Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Linguistic returns : the currency of sceptical-rhetorical theory and its stylistic inscription in the… Wittfoth, Monina 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2011_spring_wittfoth_monina.pdf [ 1.16MB ]
JSON: 24-1.0071580.json
JSON-LD: 24-1.0071580-ld.json
RDF/XML (Pretty): 24-1.0071580-rdf.xml
RDF/JSON: 24-1.0071580-rdf.json
Turtle: 24-1.0071580-turtle.txt
N-Triples: 24-1.0071580-rdf-ntriples.txt
Original Record: 24-1.0071580-source.json
Full Text

Full Text

Linguistic Returns: the currency of sceptical-rhetorical theory and its stylistic inscription in the Platonic and Derridian text  by  Monina Wittfoth B.A., The University of Toronto, 2000 M.A., The University of British Columbia, 2003  A DISSERTATION SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate Studies (English)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2010  © Monina Wittfoth, 2010  ii  Abstract Based on the premise that modernity’s understanding of the linguistic sign has a long history dating back to ancient Greece when the linguistic mediation of knowledge preoccupied thinkers like Parmenides and Plato, this dissertation synthesizes contemporary post-structuralist and rhetorical understandings of language with like-minded findings of other fields of language study. It sees post-structuralist and deconstructive understandings of language as being congruent with the long tradition of rhetorical theory and the infamous linguistic turn in philosophy, that was initiated by the Tractatus Logico-Philosophicus, as a turn away from the actual phenomena of language towards an idealization of it. Nevertheless the thesis discovers recent findings by some of the beneficiaries of the “philosophy of language” that corroborate rhetorical theory’s insights. Inspired by both Derrida and Plato, this dissertation presents a rhetorical-deconstructive image of language that, recalling the root of the term skopevw (‘I look,’ ‘behold,’ ‘contemplate’), I characterize as sceptical. This study follows the theoretical matrix of de Man, Fish, Culler and Barbara Johnson, who are, of course, themselves following Derrida. It has a holistic attitude to language characteristic of the Bakhtin/Volsohinov approach, drawing insights from classical and contemporary rhetorical theory, post-structuralist theory, findings of systemic functional grammar, recent work in cognitive science and psychology on affect and language use, the scholarship of reported speech, and the ostensive-inferential theory of communication called Relevance. This cross border work in intellectual history, language theory and stylistics examines the interstices of theory and style in the work of figures such as Plato, Aristotle, Bacon, Nietzsche, Wittgenstein, Voloshinov, Bakhtin and Derrida. Sperber and Wilson, Scholars of Reported Speech and Halliday are central to its findings; the thinking of William James and Silvan Tomkins play supporting roles. It positions Plato as the founder of rhetorical theory and studies the voices of Plato and Derrida as language theorists. I examine both how Plato and Derrida talk about language and what theories of language underlie their styles, determining finally that their sceptical-rhetorical theories prompt their ironic-conversational stylistics.  iii  Table of Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Table of Contents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v PART 1 1.1 1.2 1.3 1.4 1.5  paving the way.. . . . . . . . . . . . . . . . . . . . . . . . . . . . pretexts: Bacon, resistance and the style myth . . the road ahead. . . . . . . . . . . . . . . . . . . . . . . . . . . . W ittgenstein & linguistic complexity . . . . . . . . . . the participants and other parts. . . . . . . . . . . . . . .  ... ... ... ... ...  . . . . .  . . . . .  .... .... .... .... ....  . . . . .  .................... .................... .................... .................... ....................  .2 .7 16 19 24  2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8  adventures of signs in human communication . . . . . . . . . . . . . . . . Relevance Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . historical context: Bacon and the “plain style”. . . . . . . . . . . . . . . . epistemological quagmire: affective rationality . . . . . . . . . . . . . . . . epistemological quagmire: analogy & knowledge. . . . . . . . . . . . . . . fuzzy logic and the biochemical image of grammar. . . . . . . . . . . . . epistemological quagmire: part-whole relations. . . . . . . . . . . . . . . the inherent resistance to theory: de Man and rhetorical theory.  3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11  conn-texts & citation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 legacies: code-communication and the written language bias. . . . . . . . . . . . . . . . . . . . . . 86 legacies: performatives, seriousness & ordinary language. . . . . . . . . . . . . . . . . . . . . . . . . 89 intersections: language theory and reported speech. . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 scholarship of reported speech. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 verbatim’s fraudulent celebrity status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 involvement strategies & affective interests in speech. . . . . . . . . . . . . . . . . . . . . . . . . . 102 from involvement to performativity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Voloshinov’s sign: dancing in the caverns of our minds. . . . . . . . . . . . . . . . . . . . . . . . . 110 Nietzsche’s metaphorical turn. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 language’s bewildering complexity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121  4.1 4.2 4.3 4.4 4.5  stylistic turns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . the vicissitudes of “plain” style myths. . . . . . . . . . . . . Aristotle’s clarion call for clarity . . . . . . . . . . . . . . . . . good writing and difficult texts: clarity and opacity. . . genre expectations. . . . . . . . . . . . . . . . . . . . . . . . . . . .  ........................ ........................ ........................ ........................ ........................  124 129 137 148 152  5.1 5.2 5.3 5.4 5.5 5.6 5.7  “Platonism’s” voice-overs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . genre and the style of “propositional knowledge”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . genre and social relations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “Always with irony” (D 67). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . themes, theses, and “the ability to follow the given thread” (D 63). . . . . . . . . . . . . . . weaving, meaning, and le vouloir-dire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . reading Derrida writing about Plato . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  157 163 169 175 178 188 193  PART 2 ................. ................. ................. ................. ................. ................. ................. .................  28 31 38 42 52 57 65 71  PART 3  PART 4 . . . . .  . . . . .  PART 5  iv 5.8 5.9 5.10 5.11 5.12  Derrida’s reader and the addressee of “Plato’s Pharmacy”. Poetry’s banishment re-visited. . . . . . . . . . . . . . . . . . . . . . . . some polemical mechanics of style . . . . . . . . . . . . . . . . . . . . self-referential text as discourse about discourse. . . . . . . . . linguistic cues, ‘how things are said’ & meta-rhetoric. . . . . .  ... ... ... ... ...  .................. .................. .................. .................. ..................  197 202 210 213 220  Post Script. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237 W orks Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239  v  Acknowledgements I would like to extend my sincere thanks to a number of friends and colleagues without whom this dissertation would not have been completed. I thank Julian Patrick and Francis Sparshott who got me started on this project way back in 1983. I am grateful to Iain Higgins, who ‘gave me a chance,’ to Nan Johnson who had confidence in me, and to Adam Frank who challenged me. I owe very special thanks to my supervisor at UBC, Janet Giltrow, for her guidance, intellectual engagement and friendship on this great adventure. I would also like to thank Gregory Renault for his early support of my academic aspirations and the ‘social-theory crowd’ who guided my first steps. I am very grateful to my family and Yukon friends for their faith and unwavering support, and to Wayne and Evelyn for lending their eyes to the task of editing.  1  The notion that a theory might be extravagant is probably more controversial; but I have always seen this as a positive feature in the sense that there are more conceptual resources available than are necessitated for any particular task. This seems, perhaps, to go against the usual demand for parsimony, for “the simplest solution that is compatible with the facts”; and it is true that I see no great virtue in simplicity – I prefer the criterion of “the best tool for the job.” M.A.K. Halliday “Linguistics as Metaphor” (OL 265)  Our discussion will be adequate if it has as much clearness as the subject-matter admits of, for precision is not to be sought for alike in all discussion . . . We must be content, then, in speaking of such subjects and with such premises to indicate the truth roughly and in outline, and in speaking about things which are only for the most part true and with premises of the same kind to reach conclusions that are no better. Aristotle Nicomachean Ethics (W.D Ross 1094b).  2  PART 1 1.1  paving the way ~  Inspired by both Derrida and Plato, this literary-language theory dissertation presents a rhetorical-deconstructive image of language that follows the theoretical matrix of Paul de Man, Stanley Fish, Jonathan Culler and Barbara Johnson, who, of course, are themselves following Derrida. However, despite Derrida’s intimation that the question of the sign is “an unexpectedly historical one,”1 and despite Nietzsche’s Lectures on Classical Rhetoric,the long history of this supposedly contemporary and post-structuralist understanding of language is not fully appreciated. Plato’s treatment of language – which is especially attentive to the epistemologically problematic but nevertheless creatively productive power of tropes and affect in language use – indexes the linguistic consciousness of his audience. Michael Naas’ account of persuasion (in Turning: From Persuasion to Philosophy) and the dance between trevpw (I turn) and peivqw (I persuade), together with its suggestive middle voice (I obey), outlines a movement which signals not only the human-inter-subjective element of linguistic interactions, but also the tropological nature of language’s reality. The linguistic trope both displaces human experience and filters and shapes our account of it. The trovpoV, or ‘turn,’ is a way, path, or direction, and a manner, fashion, or custom. It is both form and content, style and meaning, both the mechanism and the goal of persuasion, a bifurcated functionality emblematic of the linguistic complexity thematized in this thesis. The contours of the rhetorical approach to language that Naas traces in Homer’s Illiad present an image of language which may begin to seem oddly modern. The work of Naas (in philosophy), Schiappa and others, like Covino (working in the history of rhetoric) have contributed to revised histories of the birth of rhetorical thought. Rhetoric’s origins are no longer figured as American rhetorical theory has wished: neither as a response to the burgeoning democracies of antiquity,  1 “The question of the sign is itself more or less, or in any event something other, than a sign of the times. To dream of reducing it to a sign of the times is to dream of violence. Especially when this question, an unexpectedly historical one, approaches the point at which the simple significative nature of language appears rather uncertain, partial, or inessential. It will be granted readily that the analogy between the structuralist obsession and the anxiety of language is not a chance one” (WD 34). See the first section on “Ancient Rhetoric” in Friedrich Nietzsche on Rhetoric and Language.  3  and not probably arising from the need to speak in the polis at all, but rather as part of the epistemological project of understanding the world that characterized the intellectual climate of Ancient Greece.2 In the days of Parmenides and Plato, when it was assumed that thinking entailed the use of language, language came under scrutiny as the medium of thought. Both the literariness and the insistent scrutiny of linguistic mediation in Plato’s dialogues have always interested me. The view of language inscribed throughout the Platonic text seemed to run completely counter to the understanding of language traditionally ascribed to “Platonism.”3 Schiappa’s conclusion that Plato coined the term rJhtorikhv (the art of rhetoric) prompts my consideration of Plato as the founder of rhetorical theory: not because he was advocating a new theoretical art of rhetoric, but because he coins this term to prompt discussion of language’s rhetorical nature, and because he explicitly analyzes the nature of the sign and takes great pains to show over and over again how it works (and doesn’t work).4 He is, moreover, clearly not discovering something about language that was not already known. From such a vantage Aristotle’s Rhetoric appears to be the codification of a centuries-old understanding of language.5 The fact then that Nietzsche’s Lectures on Ancient Rhetoric reveal a proto-post-structuralist understanding of language which has a lot in common with that of Plato and others becomes less surprising than it might otherwise seem. I wonder if modernity’s  2  See for example Leff and Welch on mis-readings of Aristotle, and Welch and Jarratt on revisions; compare with standard views in Kennedy; and note that Burke didn’t need to “extend” classical rhetoric nearly so much as he thought. 3  Cornford remarkably concludes that the “underlying assumption” of Platonic “doctrine” is that “every common name must have a fixed meaning, which we think of when we hear the name spoken” (9). However, Gonzalez’s account of the history of Platonic scholarship suggests that this “tradition” likely has its start in the eighteenth century with the end of Neoplatonism Third (viii). 4 Arthos’ investigation of “the distant ancient meaning of rhetoric” that Gadamer discusses at the end of his career speaks to this rhetorical affiliation with Plato and the epistemological focus of antiquity. Arthos observes that the “kairoticperformative rhetoric championed by Isocrates and Cicero . . . is for Gadamer . . . anchored squarely in Plato’s dialogic example” (172). Arthos argues that “the constraint upon access to knowledge disciplines metaphysics rather than rhetoric, and reorients the tradition to the modesty of its ontological competence” (172). This lines up with de Man’s account of Nietzsche’s critique of metaphysics. 5  The contradictions in Aristotle’s linguistic discussions may attest to an interest in establishing a scientific approach to language despite its obviously rhetorical nature. De Rijk’s opus Aristotle: Semantics and Ontology comments on the relationship of language to thinking in Aristotle and Plato. “Considering the fact that authors like Plato and Aristotle believed language to be the key to unravelling the mysteries of our surroundings, we should approach them in a way that reflects this attitude; thus whatever they have to say about what they consider to be the case should be examined in the broader context of Language and Reality. The semantics indeed will turn out to be the mechanism of his way of thinking and explaining things, and should therefore be used as the key par excellence to understanding Aristotle’s thought” (10). Frede and Inwood’s collection highlights the special attention that was given to language because of “the interdependence between language and thought” and “the importance attributed to rhetoric and other forms of self-expression” (4).  4  convictions about progress and advancement have blinded it to the insights of previous eras. The similar strains in the Derridian and Platonic approaches to language have guided my language-theory investigations; but they have also profited from beneficiaries of analytic philosophy whose empirical, corpus-based work (although sometimes starting from code assumptions about language) led to conclusions that supported my post-structuralist image of language. My original intention to compile a brief renewed-rhetorical image of language as a lens for analyzing both the language theories and writing styles of Plato, Derrida, Burke and perhaps even Habermas anticipated a relationship between the theory of language and the theorist’s style. Austin-Grice beneficiaries like Sperber and Wilson are very helpful in this language-theory-style project; and the gesture of bringing these formerly duelling factions together seemed productive. The scholarship of reported speech was also a natural fit with Plato’s dialogues and the enigmatic and playful approach to other voices in the Derridian text. But rather than simply supporting my work with stylistic tools, these linguistic-pragmatics fields expanded its theoretical account which contracted its stylistic analysis. The swelling of the language theory section, which seems to index the fertility of bringing these diverse voices together, reduced stylistic analyses to the Platonic and Derridian texts. In addition, M.A.K. Halliday’s essay-type theoretical summaries provided me with a new and exciting image of grammar, as well as shedding light on some of the tensions within “mainstream” linguistics. To my surprise Halliday’s summarizing essays addressed in one way or another most of the main points I was advancing about how language works. Halliday’s grammatical writing had never inspired any eureka-moments for me before I encountered these theoretical essays; my experience reading his (albeit useful) cohesion analysis had been enervating.6 Nevertheless Halliday was considered by many to be the grammar guru of the latetwentieth and early-twenty-first centuries. He is the leading proponent of the now widely recognized systemic-functional grammar, although he represents himself as working within a broad network of grammarians with varied, open-ended systemic approaches. It is significant  6  See Halliday’s Introduction to Functional Grammar.  5  for current purposes that as late as 1997 he still modestly positions against “mainstream linguistics” (OL 266). Halliday’s work endorsed my project’s speculative literary and philosophical perspective; this endorsement of the rhetorical-deconstructive image of language by a grammarian summarizing the theoretical implications of a lifetime of fine-grained, empirical studies of language at the grammatical level, calibrates an alignment of systemicfunctional grammar, ancient linguistic insights, post-structuralist discourses, and the findings of postivistically oriented research into communication and reported speech. Halliday’s theoretical summaries cast a new light on an affiliation I had been making between “mainstream” linguistics and analytic philosophy, as well as qualifying my references to the term grammar. It is the mainstream designation here that is most problematic. Halliday takes Ellis’1993 Language, Thought and Logic to task precisely for treating the entire discipline of linguistics as being aligned with analytic philosophy’s rule-governed-grammar approach. Ellis mistakenly accuses “linguistics” of code-metaphor assumptions that systemic-functional grammar repudiates (OL 232-47). However, Ellis’s failure to recognize the work of systemic grammarians attests to the perplexity and diversity in language theory to which he himself is reacting. Arguing that language theory is the purview of the two disciplines philosophy and linguistics, Ellis believes that the fracturing of these into multiple fields has resulted in a reinventing-the-wheel tendency – to “begin again at the beginning” or “revolutionize” language theory (10-11). He complains that scholars working in these fields fail “to see what the history of linguistic theory had to say” about the assumptions on which their arguments are premised (5).7 Ellis would doubtless not make the mistake of overlooking systemic grammar now. Although Ellis and I have radically different approaches, the tenacious grip that the code metaphor has exerted in some fields of language theory over the past few centuries and the failure to see what was already known about language that he complains about has propelled my text as well. Halliday’s acknowledgement of Whorf’s great influence on his thinking is corroborated 7  John M. Ellis remarks that language theory is characterized by inconsistencies, amnesia, and a lack of rigour when it comes to dealing with issues that were already problematized by Wittgenstein and Saussure.  6  by Halliday’s work, which evinces the decidedly post-structuralist comportment that Whorf and Sapir have bequeathed to their language-theory legatees (OL 188). The contrast between the Whorf-and-Sapir orientation and the analytic-philosophy approach lines up with the distinction that Voloshinov makes between “two trends in linguistics”; Voloshinov tracks strains of linguistics amenable to post-structuralism back to Herder in the eighteenth century. Voloshinov’s albeit crude and repeatedly qualified distinction is a useful heuristic for gauging the positivist and post-structuralist orientations in linguistics. He defines the first trend, “individualist subjectivism,” with reference to the comparative-descriptive linguist Wihelm von Humboldt, who, he claims, was influenced by Johan Gottfried Herder (48). Interestingly, Herder’s belief that “thought and language are inseparable” carries on the thought-language ligature from antiquity that is frequently referenced in this dissertation (Honderich 352). Overall, the holistic and interdisciplinary approach of Voloshinov’s first trend is consistent with the approaches of both this thesis and Halliday’s summaries. Humboldt imagines language as a fluid, social and creative process, depending on laws of individual psychology (Voloshinov 49). Voloshinov attributes the second trend, “abstract objectivism,” to the rationalism of the seventeenth and eighteenth centuries, noting that “the ideas behind the second trend received their first . . . expression in Leibniz’s conception of universal Grammar” (57). He says “the second trend has profound inner connection with Cartesian thought and with the overall world view of neoclassicism and its cult of autonomous, rational, fixed form”; and he claims that “the idea of language as a system of conventional arbitrary signs of a fundamentally rational nature was propounded by representatives of the Age of the Enlightenment in the 18th century” (57-58). Notwithstanding his own obvious admiration for Saussure and despite Saussure’s widespread influence both within linguistics and outside of it, Voloshinov identifies him with this abstractobjectivist trend. Saussure is exemplary here for having defined his “object of study” by isolating the linguistic sign from its social and psychological roots. Without diminishing the fruits of Saussure’s lucid analyzes, his gesture severs the linguistic sign from the life forces that  7  create and shape it.8 Voloshinov perceives a genealogical linkage between Leibnizian universal grammar, conceptions of autonomous, rational, fixed forms, and rule-governed grammar. Rulegoverned grammar – a set of logical, learnable rules that governs how language works and how we understand one another – wants the truth conditions of propositions to determine meaning. This rigid deterministic conception of grammar is radically different from the approach for which Halliday is the exemplar here. In Part 2, I will outline Halliday’s biochemical image of grammar and attempt to distinguish the (increasingly mainstream) grammarians and linguists, like Halliday, from those adhering to the rule-governed grammar that Leibniz imagines. This thesis somewhat arbitrarily refers to “mainstream linguistics” as excluding the important field of systemic grammar in order to distinguish analytic philosophy’s beneficiaries from those of Herder, Humbolt, Whorf and Sapir. Although I suspect that Halliday would side with Kenneth Burke in framing himself as a constructionist rather than a deconstructionist, Halliday nevertheless gets at many of the same kinds of issues about language that Derrida does, making him an excellent complement to the other language theorists convened in this dissertation.9 1.2  pretexts: Bacon, resistance and the style myth ~  Francis Bacon has long and misleadingly been revered for having inaugurated the “style of science.” This style of writing, originally characterized as “plain,” has since Bacon’s time morphed into modern conceptions of simplicity, clarity and ultimately transparency: terms that not only mean entirely different things, but that are clustered together as though naturally affiliated and proclaimed as the writing ideal.10 Bacon’s style-of-science affiliation provides an opportunity to foreground two themes of interest to this thesis. On the one hand, scholars studying research writing, like Giltrow, find that this simple, clear style of science simply does  8 Similarly, Halliday’s esteem for Saussure does not prevent him from remarking the problem of separating langue and parole, and from asserting the sign’s lack of autonomy – a matter which Saussure clearly recognizes (OL 267, Saussure 117). 9  Halliday, for example, says that “language construes experience in terms of complementarities” (citing as an example terms which are both “bounded” and “unbounded”). For Halliday this means that language constructs reality in terms of binaries; and he considers this contradictory-binary richness part of the “extravagance of language” (OL 265). Not only is Derrida famously fascinated with language’s dependence on differentiation and binary pairs (like form and content) and the privilege ceded to one side of an opposition (like speech and writing), Halliday’s “extravagance” is very like the exorbitance or supplementariness that Derrida treats with such insistence in Of Grammatology and elsewhere. 10  See Zappen 1989 and Lipson documentation of references linking Bacon to this style.  8  not exist. Contrary to this imagined style of science, linguistic-pragmatics studies of research writing show that, completely apart from such obvious matters as register (or specialized lexicon), the research genres are strongly marked by disciplinary-ethos investments; they have rhetorical aims directed at such matters as disciplinary-knowledge construction and credibility, displays of respect, identification, and community-solidarity building; and they are addressed to specific and limited audiences – hence hardly “plain” or transparent to non-specialists (e.g., Hyland, Meyers Politeness, Swales, et al.). Giltrow concludes that “the language of science,” which is “seeded with social interests” designed to convey community credibility and the stance of neutrality, is anything but unstyled (Modern 173).11 The supposed transparent style of science is a myth: but a powerful value-shaping myth, and important here not only because of the prestige it claims for itself but because of the understanding of language it promotes. On the other hand, as far as Bacon himself is concerned, recent research in the history and rhetoric of science reveals that his views are not so straightforward. I believe that thinking of him in terms of this plain-style-of-science myth reduces and trivializes the linguistically insightful comments scattered throughout his writings. But the fact of his having been taken as Mr. Scientific Plain Style is evidence or measure of the force and currency of problematic conceptions of linguistic transparency, which wrongly affiliate plain, clear, simple and true.12 For example, consider the implications about language from the following passage from Novum Organum: But the Idols of the Market Place are the most troublesome of all – idols which have crept into the understanding through the alliances of words and names. For men believe that their reason governs words; but it is also true that words react on the understanding; and this it is that has rendered philosophy and the sciences sophistical and inactive. Now words, being commonly framed and applied according to the capacity of the vulgar, follow those lines of division which are most obvious to the vulgar understanding. And whenever an understanding of greater acuteness or a more diligent observation would alter those lines to suit the true divisions of nature, words stand in the way and resist the change. Whence it comes to pass that the high and formal discussions 11  These scholars analyze a range of disciplinary research genre in science (like math and cell biology), social science and humanities disciplines. Giltrow, Hyland and Myers have analyzed in particular writing in the so called “hard sciences.” 12  An application of Halliday’s cohesion analysis to these collocates would doubtless be a fruitful index of Enlightenment thinking about language. See Cohesion in English.  9  of learned men end oftentimes in disputes about words and names; with which (according to the use and wisdom of the mathematicians) it would be more prudent to begin, and so by means of definitions reduce them to order. Yet even definitions cannot cure this evil in dealing with natural and material things, since the definitions themselves consist of words, and those words beget others. So that it is necessary to recur to individual instances, and those indue series and order, as I shall say presently when I come to the method and scheme for the formation of notions and axioms. (emphasis added LIX)13 Like other thinkers from Plato to Derrida, Bacon quite typically comments here on such matters as 1) linguistic mediation, the shaping influence of words on human perception (“words react on the understanding”), 2) genre, register and audience design, the different meanings that words have in different types of discourse with different addressees (words that might be “framed and applied according to the capacity of the vulgar”), and 3) the problem of linguistic recursion mise en abîme (the open chain of signification). Bacon’s allusion to the troublesome and perpetual need to define words both echos Plato’s Cratylus14 and anticipates the Derridian critique: “definitions themselves consist of words, and those words beget [other words].” The disparity between Bacon’s rhetorical understanding of language and positivist assumptions underlying the plain style reveals certain language-theory misconceptions of interest to this thesis. Both Bacon’s views of language, and the fact that he is remembered for an opposite view, make him an exemplary site for introducing some of the features of language that are now associated with post-structuralism – but which actually have a history dating back to ancient sceptical views of language.15 The figuring of Bacon as Mr. Scientific Plain Style is paradigmatic of certain language-theory misconceptions that de Man explores in “Resistance to  13  Throughout this dissertation if emphasis added to citations it is noted parenthetically, as here.  14  While it is in the context of Cratylus that Sallis comments that “it begins to become evident to what degree the entire enterprise is bound up in names, explaining names by other names, which themselves would always require to be explained by still other names, and so on without limit” (257), this discussion also appears elsewhere. See for example, Sophist 244d. 15 The term sceptical can be taken as an adjective here, but the connections to ancient scepticism also exist, a matter to which we return. For the purposes of this thesis I will treat Sextus Empiricus’ (160-210 C.E.) comments as exemplary. Leo Groarke claims that Sextus Emipiricus “describes the ‘skeptic’ (from a Greek verb meaning ‘to examine carefully’) as an ‘investigator’ (a ‘zetetic’). According to Sextus, the skeptic is someone who has investigated the questions of philosophy but has ‘suspended judgment’ (practising epochê) because he is unable to resolve the differences among the contrary attitudes, opinions and arguments he found. Instead of adhering to a definite philosophical position, the skeptic is someone who continues to investigate.” Sextus particularly foregrounds “the skeptic’s open-minded attitude to the possibility of apprehending truth” – however at the same time sceptical arguments “raise deep questions about any claim to truth.”  10  Theory.” Such uptake failures of rhetorical-sceptical understandings of language, which seem to gather critical mass at various historical moments of scientific over-exuberance, are of central concern to this thesis. Hence Bacon’s writing about writing and de Man’s resistance article offer useful opportunities (or sites) for raising certain issues about language and as a basis for getting started on this dissertation’s inquiry into language theory. I am borrowing classical rhetoric’s enigmatic metaphor tovpoi (‘places’) in order to describe the way that Bacon’s and de Man’s discussions about language and the plain-style myth itself are sites (or places) that carry with them useful contextualizing resources. I am using Bacon, the resistance metaphor, and the plain style as specific topoi (‘places’) for my discussion; exploiting them opportunistically, I use them as a pretext for discussing ancient linguistic insights, theory-uptake failures, problematic linguistic intuitions, Cratylian conceptions, code and conduit metaphors, etc.16 According to de Man such discourses about language (as Bacon’s above) that theorize about linguistic signs engender a “resistance to theory” which he claims, “is a resistance to language about language” (13). (In this vein, we shall later see the way that talk-about-talk in the Platonic text has been largely overlooked.) Resistance to language theory has its roots in fantasies about human knowledge, as we shall see. It is also a resistance to thinking-aboutlanguage that arises from unexamined intuitions of linguistic transparency. De Man says, “it is a resistance to language itself [resistance to language’s tropological nature, for example] or to the possibility that language contains factors or functions that cannot be reduced to intuition” (13). Since language use seems intuitive, we assume de Man says “that when we refer to something called ‘language,’ we know what it is we are talking about” (13); and we seem to assume that we know how it works (without thinking about it). But intuitions are sometimes wrong. As we shall see, Sperber and Wilson’s inferential model of communication together with the linguisticpragmatics scholarship on reported speech shows that only some uses of language are descriptive; and reference (a sign-to-thing correspondence) often plays an amorphous and 16  Aristotle’s common topoi are generally understood as lines of argument, known commonplaces that are recognizable patterns of thought exploited by analogy to a matter at hand. While McKeon considers all topoi to be place of invention, Aristotle’s specific topoi are a subject of considerable debate. See McKeon Rhetoric, C. Miller Special Topics, Leff Topics, Leff Topoi, McAdon Signs.  11  whimsical role. Completely apart from the problem of intuition’s empirical reliability, there is some question of whether “intuitions” about language are actually “natural” or whether they might be schooled and historical. The sender-to-receiver-transmission-of-information codemetaphor (kin to transparency’s conduit metaphor) may still be “intuitive” for some; but we shall see that, contrary to the hunches we might have, some operations of language work in surprising ways. This thesis aspires to complicate rather than simplify images of language, arguing for a linguistic complexity that defies being definitively codified. My dissertation aspires to promote a linguistic consciousness that defies being tied down into “a single guiding thesis,” as Derrida will say (D 7). It is on the basis of what he calls Hegel’s Cratylian conception of language – in which meaning is essentially embodied in names, names which, like onomatopoeic words, signify the essence of the thing named – that de Man indicts Hegel in “Resistance to Theory.”17 The Cratylian conception imagines a language in which “the phenomenality of the signifier, as sound, is unquestionably involved in the [referential] correspondence between the name and the thing named” (10); words denote, or refer to, things by some kind of essential tie between them, and the meaning of a term is anchored by the essence of the thing. But de Man points out that the “phenomenal” attribution is a mistake; “the [referential] relationship between the word and the thing is not phenomenal but conventional” (10).18 It is old news that both word and thing are phenomenal (that is, they are sound, text, event, concept, object), and that their referential relation to each other is conventional. De Man’s concern is to underscore that it is a mistake “to confuse the materiality of the signifier [sound/text/image] with the materiality of what it signifies [concept/thing]” – a mistake that is frequently made (11). De Man observes that the “conventional” relationship between word and thing gives language “considerable freedom from  17  This is the conception of language parodied by Plato’s dialogue, Cratylus, named after the character in the dialogue who attempts unsuccessfully to defend this view. 18 De Man’s term conventional, which somewhat correspond’s to Plato’s use of oJmologiva (agreement) in the Cratylus, is an alternative to Saussure’s term arbitrary. Derrida aptly observes that the sign’s “arbitrariness” is “grossly misnamed” (OG 44). He argues that only the structural “relationships between specific signifiers and specific signifieds would be regulated by arbitrariness” (OG 44). He suggests that there is a “discontinuity” in the structure, not the historical origin of signs (OG 326). Signs have a history; a sign’s meanings are governed by its history not arbitrariness.  12  referential restraint” (10). This “freedom from referential restraint” is not a complete absence of referential correspondence; it is a kind of referential promiscuity. More importantly, the problem de Man is concerned with is not language’s very real referential and descriptive functions, which to judge by their continuous service are relatively effective. Rather, twentieth-century philosophy of language’s and “mainstream” linguistics’ mistaken understandings of the status and roles of reference and description in language use are at issue. A confident attitude towards the singular importance and role of descriptive referentiality in language use cohabits with the codemetaphor intuition. Sperber and Wilson describe this sender-to-receiver-signal-transmission idea as “packing a content into words” and “sending it off to be unpacked by a recipient at the other end” (1). When the Chomskian-linguist Pinker, for instance, asks “what is the trick behind our ability to fill one another’s heads with so many different ideas?” – he makes it sound as though the idea were popping out of one mind, then scurrying over and jumping into another (1). The one-way transmission of information like a package is delivered to the hearer by the speaker; the speaker thus informs the hearer with a pre-packaged, descriptive message-content. Halliday flatly dismisses the code metaphor and the paramount status of reference: when language is reduced to the status of an (imperfect) formal system, a “code” to be measured by its correspondence (or lack of it) to some supposedly independent reality, whether material or mental, by reference to which its meanings are judged to be (or else not to be) “true,” the resulting vision is so impoverished that serious questions about language can hardly even be raised . . . (OL 237) The code metaphor depends on an essentialist, Cratylian conception of language, in which meaning inheres in the word itself – a conception that confounds our understanding of how language actually works and that Halliday thinks precludes serious questions about language. Although the de Manian account of language’s referential promiscuity may seem counterintuitive, it is consonant not only with Halliday’s perspective but also, as we have noted, with findings of linguistics and pragmatics scholars. Pyscholinguists Clark and Gerrig observe that although “most theories of language use take for granted that all language use is  13  description,” reported speech is a type of “demonstration” which is not descriptive (764). These “demonstrative” functions of language (explored in detail in Part 3) push description and reference to the back seat, and are, I think, more aptly described by the contemporary expanded use of Austin’s term performative.19 In Halliday’s four metafunctions that define the dimensions of semantic space (experiential, interpersonal, logical and textual) it is the experiential and interpersonal functions that hold centre stage. He claims that the work of grammar is both to “construe” experience (by which he means construct) and to “enact” the interpersonal (Matter 63). Both of these metafunctional operations of grammar pertain to how language acts in the world that I consider the performative dimensions of language. De Man’s optimistic sounding “freedom from referential restraint” (or referential promiscuity) has the worrisome implication that language is epistemologically “suspect,” and its use no longer “determined by considerations of truth and falsehood” (10). These truth-status considerations pertain to language’s misunderstood relationship to logic (and thus to rulegoverned grammar) as we shall see. Ellis calls this the “verification” theory of meaning based on the (Cratylian) belief that “words labelled things” that they “denoted” or to which they “referred” (4). De Man’s lighthearted characterization of referential unreliability as freedom playfully mocks the epistemological anxieties that might result when the reliability of the linguistic medium is undermined. Since the epistemological functions of language depend on description and reference, when language becomes epistemologically suspect our understanding of the world is at stake. For some this news might be disturbing and resistance-provoking. But such linguistic-mediation worries are an old story, dating back to ancient scepticism and stemming from the view of language evident in the incipient rhetoricity of the Homeric text.20 From a sceptical viewpoint our cognitive dependance on language together with language’s  19 20  This thesis uses the term “performative” in the broad sense of the term following Culler (1997).  See Richard Bett (Pyrrhonism), Cherniss, and Román Alcalá on the link between Plato and ancient scepticism, and see Naas on Homer’s rhetoricity. The various shades and schools of ancient and modern scepticism that are classified in ancient philosophy are not especially relevant to this discussion. My references to scepticism are to a “loose sense” (as Bett says in Nietzsche) of suspending judgement in one’s investigation – which importantly presupposes that one investigates.  14  fundamental metaphoricity21 make all human understanding linguistically mediated and flawed (or at least highly selective) as a result. Halliday doesn’t even acknowledge a purely referential linguistic ideal when he gleefully embraces language’s “semogenic power” – its meaningmaking, construing-constructing-enacting effects (Matter 63). For sceptical language theorists the selective and shaping influence of language on human understanding (linguistic mediation) does not mean the sky is falling; it does not imply complete blindness; it simply means that humans are somewhat myopic and do not have X-ray vision. Language users need to take their vision limitations into account.22 Richard Bett’s investigation of Nietzsche’s relation to ancient scepticism observes that “Nietzsche always dissented from a certain traditional optimistic picture of the prospects for philosophy, or for inquiry more generally.” Bett describes the “traditional” picture as faith that the world is as it is objectively – that is, independently of our attempts to describe it; yet our accounts of how the world is have a real possibility of being objectively correct – that is, of describing the world as it objectively is – and we have a real possibility, when this is so, of knowing that it is so (Nietzsche 68). This “traditional” picture (which may not have begun in earnest until the Enlightenment) underwrites the Cratylian conception of language that this thesis tries to displace. Arguably, Thomas Kuhn has disposed of the fantasy that we will necessarily know when we have accurately identified something, and postmodernity has lost faith in the human potential for objectivity. How can we objectively see something like language that we are so intimately involved with? Ann Shukman characterizes the philosophy of language propounded by the French post-structuralists as coinciding with “a general skeptical attitude to language to be found in Wittgenstein and most modern thinkers since him” (563-64). It is the investigative and open-minded implications of the term scepticism associated with Sextus Empiricus’ definition of it that I wish to exploit here, and not some metaphysical 21 Language’s metaphoricity – its fundamentally figurative or, as de Man calls it, tropological character – stems from its semiotic nature; signs are conventional phenomena (not essential emanations of things as Cratylians wish); rather they are images of images of things (and actions, relations, etc.), as we shall see in the discussion of Nietzsche. 22 It is this shaping influence of linguistic mediation that prompts Halliday’s comments on neutrality and ideological inscription in “Is the Grammar Neutral?” Halliday suggests that a degree of deconstruction is always required in order to assess the perspective from which an observation is made (OL 286).  15  nonsense that nothing can be known (whatever that means), nor reform positivism that wants to solve language’s shortcomings by remaking it.23 Recalling the root of the term sceptical – skopevw (I look, behold, contemplate) – I characterize as sceptical certain theorists who attend to and investigate language instead of trying to systematize it, scholars who study semiotic material instead of trying to idealize it. These are thinkers who, despite the non-ideal character of their subject matter and undaunted by the epistemological implications of language’s nonideality, often treat it with optimism (like Bakhtin and Halliday), with humour (like Plato), and even a perverse glee (like Nietzsche and Derrida). O’Leary-Hawthorne comments on the “linguistic scepticism that runs through the [Platonic] dialogues” (168), but this sceptical view of language does not preclude investigation. As Sallis reminds us, Plato unendingly exhorts readers to use words as the medium of understanding, relentlessly advocating the pursuit of wisdom using lovgoi (‘words,’ ‘accounts,’ ‘discourses,’ ‘stories’).24 Language’s metaphoricity (which we examine through Nietzsche in Part 3) and our dependence on thinking in language and by analogy (which we examine in Part 2) means that human cognitive operations produce results of varying accuracy that are difficult to measure. And there are many characteristics of signs and their human users that contribute to language’s referential failings. Epistemological frailty that disrupts confidence in the validity of human knowledge (discussed in Part 2), self-referential writing that leaves readers uncertain about what the writer is directing our attention to (discussed in Part 5),25 and any number of language’s eccentricities and exploits (discussed throughout this thesis) might induce anxiety and thus play a role in the resistance to theory.  23  See footnote #15 on Sextus Empiricus’ definition of sceptic.  24  Gonzalez’ introduction to The Third Way re-charts the horizon of Plato scholarship in terms of sceptical and doctrinal impulses dating back to Aristotle. This remapping of the terrain is especially interesting because the post-neoplatonist shift (that occurs before the eighteenth century) lines up with Enlightenment thinking and mis-uptakes of Bacon’s views of language. 25  For example, at the end of the Protagoras after Socrates has brought Protagoras to his knees argumentatively and we are directed back to the opening scene and lines where Socrates had presented Protagoras as “the wisest of our generation” (Loeb 309d). Or did he really? Certainly on first reading that might be the impression one gets, but when we are directed back to the beginning we get a different impression. The ground seems to slip away, as the texts changes meaning. I have consulted the Hamilton Cairns Collected Dialogues, the Cooper Complete Works, the Loeb editions, Grube’s translation of the Republic and John Sallis’ own translations. Accordingly the citations of Plato will refer to the translator or to Loeb editions.  16 1.3  the road ahead ~  Did resistance, then, play a role in the repression of Bacon’s rhetorical understanding of language? What effect did the Enlightenment’s scientific super-exuberance have on conceptions of language? The Royal Society’s interests in empirical knowledge appears to have blinkered the uptake of Bacon’s sceptical views of language. Although late modernity seemingly swallowed the Royal Society’s Bacon-story whole, what is more to the point for this thesis is the reception of Plato’s, Nietzsche’s and Derrida’s discussions of language. How did the rhetorical insights of thinkers like Plato and Nietzsche influence philosophy’s “linguistic turn” in the work of Frege, Russell, Moore and the early Wittgenstein? Has the purported “plain style of science” or the mythical “style of non-style” played a role, not only in obscuring millenniaold insights into language’s operations, but also in fostering resistance to theory? This exploration of language theory synthesizes central insights from rhetoricaldeconstructive-post-structuralist lines of thought from Plato to Derrida; it shows in broad strokes how these views of language line up with findings of recent linguistic research in human-science disciplines and branches of language theory stemming from analytic philosophy. Voloshinov’s theoretical work – which melds well with not only Derrida and Halliday, but also Sperber and Wilson and the scholars studying reported speech – sets the stage for the linguistic complexity that I believe must always remain in view for fruitful, detailed analysis of language in any branch of language study. Jeanne Fahnestock’s account of how “classical and early modern attention to language” was obscured by later interests foregrounds the linguistic awareness or consciousness evident in the rhetorical and dialectical manuals of ancient and early modern times (emphasis added Cognitive 161). This attention to language (attributed to both Plato’s and Bacon’s eras) is the principle concern here. Instead of striving for a “robust theory” or “thick description” that others have called for,26 and lacking confidence in the utility of language-theory codification for non-specialists, this dissertation seeks a generalized understanding of language. Attention to language – a greater awareness or linguistic  26  e.g., Fahenstock Enriching, Fowler, and Halliday Meaning.  17  consciousness of both linguistic mediation and language’s bewildering complexity – seems warranted. Rather than advancing a new theory, I want to synthesize a common stream of related linguistic insights with a long history which have been frequently overlooked in the last few centuries. The conception of language implicit in various linguistic assumptions and attitudes is particularly interesting. Ellis, for instance, complains that language theory’s missteps often involve “assumptions built into what look like statements of the obvious” (15): hence Richard Robinson’s claim that “the purpose of a name is to refer us to a thing” (Criticism 334). Such obvious and facile assumptions are responsible for the Cratylian impulse: the word-thing fixation to which, Ellis believes, linguistics and philosophy continually revert under a variety of names (i.e., code, information-content, truth claims).27 For sceptical theorists attentive to language’s metaphoricity such recidivism is mystifying. Language’s powerful affective elements inevitably embroil analysis of it in social, psychological, and communicative dimensions; importantly, these matters are always in play in language’s epistemological uses. At this historical moment code-metaphor conceptions and faith in linguistic transparency, probably sponsored by late modernity’s overwhelming technological successes, continue to muddle understanding. If we can put a man on the moon, language must be working the way we assume it does. And language is working, just not the way we might think it does. If science, with its plain-style myth and ethos of objectivity and neutrality, is an idol of late modernity’s scientistic culture, it is unsurprising that code metaphors, determinate reference, and linguistic transparency have such a grip on conceptions of language. Pure neutrality (or objectivity) demands this Cratylian image of language.28  27  Ellis describes the code metaphor’s boomerang tendency in twentieth-century language theory: repeated acceptance of code after repeated deconstructions of it, and acceptance without any explanation for, or efforts to address the previous scholarship refuting it. He links the code metaphor to conceptions of communication and faults language theory for treating language as though its purpose were communication (15). 28 Giltrow writes that “the expert voice is known and valued for its objectivity and its neutral stance. In turn, its authority derives from this neutrality: when experts present research, their way of doing so shows that they have no personal stake in the knowledge reported” (Modern 171-72). She believes it is a stance, however, a stylistic effect and not a quality of perception. If it were imagined as a quality of perception, pure neutrality or objectivity would require Cratylian code for neutral, objective expression of what was perceived.  18  At this historical moment and in this ideological context – when code-conduit-metaphor assumptions have been eroded by post-structuralism, but nevertheless still linger in aesthetic values about writing, in assumptions about scientific knowledge, and in the reference posts used by “mainstream” linguistics – a revived and renewed image of the linguistic sign seems warranted. I am trying to displace the Cratylian-code-conduit metaphor ideas about language with an image of an affectively charged, metaphorical, semiotic material that not only mediates every aspect of human experience, but by nature both generalizes and reifies: two salient actions undertaken by linguistic signs often not taken into account. I do not intend to offer a complete picture of language, or how it works, or even of the sign itself; but by keeping socialpsychological factors of linguistic mechanisms in view, together with logical-structural ones, and by foregrounding the discrete nature of epistemological and communicative aspects of language use I hope to reinforce a sense of language’s perplexing complexity. It is not just the multi-purpose nature of language that creates complexity but that multiple purposes are served at the same time. From Halliday’s perspective there are always both construing and enacting (interpersonal) elements in an utterance at the same time. Multiple mechanisms in language – describing, identifying, constructing, communicating – operate at the same time.29 This simultaneity makes it impossible to analyze linguistic operations correctly without recognizing and accounting for multiple concurrent layers. Halliday says “one of the sources of [linguistic] complexity is that there is more than one kind of meaning in a language,” and as a result “grammar is doing more than one job at once” (emphasis added Matter 63). Borrowing a metaphor from Matthiessen, Halliday fantasizes about a “theory designed to represent the multidimensional ‘architecture’ of language” (Matter 77). It is this linguistic complexity that problematizes any effort to articulate a coherent theory of language (such as Wittgenstein attempted to do with his notes when he set out to write Philosophical Investigations), and that prompts the stylistic eccentricities of some sceptical-language-theory  29 Notwithstanding disputes about the meaning of his terms, Austin also, of course, claims that “to perform a locutionary act” is “also to perform an illocutionary act” (How 98); “in general the locutionary act as much as the illocutionary is an abstraction only: every genuine speech act is both”(147).  19  texts (i.e., Parmenides’ poem, Platonic dialogue, Nietzsche’s prophetic voice, Derrida’s metaphors, Wittgenstein’s aphorisms). Fahnestock is interested in “the view of language implicit in rhetorical stylistics” (Cognitive 165). I am interested in the way conceptions of language prompt different types of stylistics. In particular, I want to consider the way language theories are stylistically inscribed in sceptical-language-theory texts. 1.4  Wittgenstein & linguistic complexity ~  It seems likely that linguistic complexity results in a certain codification resistance as well as a certain summary resistance of stylistically eccentric texts (i.e., Plato, Derrida Wittgenstein). Linguistic complexity also figures into such anomalies as this thesis’ use of a referentially descriptive mode to refer to potential referential failure and underwrites the challenges of taking a wide-angle view of language into account. Wittgenstein’s “Preface” to Philosophical Investigations is useful here because it stages some of the conceptual and stylistic challenges of language-theory texts central to this thesis. His prefatory comments are thus a topos that introduce thematic considerations. For example, even something as trivial as his complaint that his ideas, in both oral lectures and written publications, have been “misunderstood” and “mangled” is representative. Although any writer might make this claim, as Culler and Lamb’s Just Being Difficult: Academic Writing in the Public Arena suggests, it seems a fortiori a problem for sceptical language theory.30 Such misreadings and manglings of language-theory texts haunt this thesis from its beginning with Bacon to its ending with Plato. These mis-takes seem symptomatic of a certain resistance to theory. While de Man’s resistance metaphor signals both reading failures and the misguided intuitions and anxieties of readers, what about writers? How do worries about writing sceptical language theory figure in stylistic choices? And how much does a concern with implicitly refuting their own premises – the tendency for sceptical theories of language to undermine themselves, for instance, by referring to potential referential failure – determine what language 30 See also Nietzsche’s remarks “On the question of being understandable” (GS 381) and “We incomprehensible ones. – Have we ever complained because we are misunderstood, misjudged, misidentified, slandered, misheard, and not heard?” (GS 871). See also Derrida’s engagement with charges levelled against him: “logical contradiction,” “performative contradiction,” “relativist,” “nihilist,”and a non-serious philosopher (Monolingualism 2-4). The sheer volume of texts dedicated to the problem of “how to read Platonic dialogue” speaks for itself.  20  theorists say and how they talk? Overtly performative uses of language might be more desirable than ostensibly referential-descriptive uses burdened by paradox and contradiction; concerns about academic writing’s objective-rational ethos could discourage writers from making observations whose self-referentiality is potentially undermining. I wonder how the languagetheory subject matter determines stylistic questions (e.g., genre, mode, approach) and whether the image of a descriptive-referential academic writing that disavows the kind of footloose stylistic we find in Wittgenstein and Derrida is itself hostage to a mythical style of science. This thesis is particularly attuned to how such considerations might be evident in the stylistic choices of Derrida and Plato. When questions about the style and ethos of academic writing (with its stance of neutrality) are tied together with a theorist’s desire to provoke thought, the threads get quite tangled. Wittgenstein says “I should not like my writing to spare other people the trouble of thinking. But if possible to stimulate someone to thoughts of his own” (“jemand zu eigenen Gedanken anregen”) (Preface). This desire to prompt or provoke [anregen] another’s thinking presents a reader-writer relationship with a more intimate (playful, invasive) cast than the respectful (collegial, “objective”) character31 of the reader-writer relationship in academic writing. This thesis is interested in the subjectivity of reader and writer in language theory texts. Wittgenstein’s interest in provoking reader thinking highlights his designs on his reader, warranting consideration of his relationship to his reader in terms of Sperber and Wilson’s model of ostensive-inferential communication. The salient role of reader inference that is documented by Sperber and Wilson directs attention to both the deciphering role of the reader and the social implications of leaving things implicit. Thus, for example, irony creates the impression that the writer shares a view with the reader; it is something they have in common, a kind of intimacy (242). Eliciting this sense of shared vantage is what Tannen calls an  31  Giltrow cites Biber and Finegan’s “taxonomy of stance styles” in different genres of which “faceless is the one that fits research writing. Facelessness is a ‘marked absence of all stance features’” (Modern 173). The stance of facelessness seems to be what researchers learn to do when they find their research voice within in their particular disciplinary field and they seem to do so without considering its linguistic implications. If they actually imagine themselves as neutral objective observers, they will tend to resist perceiving their linguistic usage as a stance and resist recognizing their own designs on their readers. Having designs on one’s readers clearly precludes pure neutrality and is incongruent with the ethos of neutrality.  21  “involvement strategy,” as we see in Part 3. But how “involved” are writers with readers when the stance marking their prose is neutrality? Writers’ attitudes towards readers – what readers are expected to know and recognize and how much the writer presumes to inform the reader, or engage them in play – are just a few of many ways of evaluating the profile of a text’s reader. This thesis considers the differences between reader-writer relationships in texts designed to inform, to remind, or to provoke and then explores the role of reader-writer relationships in the Derridian and Platonic texts. It tracks the kinds of implicatures Wittgenstein, Derrida and Plato expect their readers to make. The discussing-observing-concluding (rational-objective) demeanour of research writing might look askance on the provocation motive which seems too intimate, playful, and demanding (not very neutral) for academic writing. Research writing does not seem intended precisely to provoke in the arresting way that Philosophical Investigations does. Aristotle, whose text serves as primogenitor of academic writing in this thesis, advises that “what is written should generally be easy to read” (On Rhetoric 3.5). Posterior Analytics outlines a generalized method of proceeding from known to unknown because learning “proceeds from pre-existent knowledge” in the form of an already known “fact” or the “meaning of a term” (Loeb 1.1 71a).32 Fine-grained, linguistic-pragmatics analyses find that at a word-order level given information precedes new – we begin with what is known before introducing something new. Such oddities as reversing given and new, or confounding genre expectations at the particularistic, pragmatic, word-order level, can arrest reading fluidity and be off-putting to readers. Riffaterre and others suggest that stylistic anomalies catch our attention and make us stop to think. As we shall see at the end of Part 4, the expectations that bring stylistic anomalies into relief are established by genre. It seems likely that writer interests in provocation might prompt some theorists (like Derrida and Wittgenstein) to subvert stylistic norms of academic writing. Possibly a motive such as Wittgenstein’s provocation can shed light 32  Citations from Aristotle’s Rhetoric are usually from Kennedy’s On Rhetoric. I have also used Cooper’s translation of Rhetoric and the Loeb editions, as well as McKeon’s Introdcution to Aristotle, and an online text for a portion of the Metaphysics translated by W.D. Ross. I will indicate the translator or Loeb edition for all translations except On Rhetoric which is Kennedy’s unless otherwise noted.  22  on the footloose stylistic of Plato, Nietzsche, or Derrida. Wittgenstein’s realization that “the nature of the investigation” prohibited turning the Investigations into a “book” (“einem Buche”)33 is instructive here. He says “my thoughts were soon crippled if I tried to force them on in any single direction against their natural inclination. – And this was, of course, connected with the very nature of the investigation. For this compels us to travel over a wide field of thought criss-cross in every direction” (Preface). The aphoristic style of Philosophical Investigations is paradigmatic of the ways that language-theory texts perform their own theories. Wittgenstein’s “remarks,” he tells us, constitute “sketches” illustrating some feature or features of a linguistic usage (Preface). The carefully choreographed dance performed by these seemingly artless remarks both comments on and, at the same time, demonstrates the diversity of kinds, functions and purposes of “names.” For example, the same name serves as different kinds, for different functions and different purposes; he queries whether or not many words name anything at all. This protean quality of names – only one dimension of linguistic complexity, albeit an important one – adapts them to the determining role that context and genre play in linguistic operations, that is, to the extent that language is determinate at all. Wittgenstein tells us he suddenly realized that he “should publish those old thoughts and the new ones together: that the latter [Philosophical Investigations] could be seen in the right light only by contrast with and against the background of [his] old way of thinking” (Preface). Context is a dimension of language that seems to dominate all levels of linguistic analysis, from the most particularized linguistic studies of phonemes to the most generalized levels of ideological activity. Context provides a place or location which situates linguistic interpretation – whether the place is a generalized, high-level argumentative structure or a very particular scene with specific discursive resources. Context has many faces. It can be viewed through many levels of magnification and produced by a variety of rhetorical-stylistic gestures. In  33 Note also Derrida’s opening line to Dissemination: “This (therefore) will not have been a book”(3). We will return to the question of writing a coherent theory of language in the context of considering the implications of totalizing texts and Bacon’s admiration of aphorism.  23  Philosophical Investigations, the need for context underwrites the use of the Tractatus as the background on which the sketches are drawn. An assumption, however apt or erroneous, from the earlier work both prompts a sketch and then becomes its context. Here what is contextualized are whole, historically-situated discourses, disciplinary lexicons, and approaches to language. Wittgenstein’s often Cratylian interlocutor presents these attitudes, summoning various positions that were pursued in early-twentieth-century philosophy of language. However, the voice of Wittgenstein’s fictional interlocutor together with the arrangement of counter-punctual remarks are stylistically distinct from the way that Sperber and Wilson, for example, contextualize their views. Wittgenstein’s style of contextualization reflects the linguistic complexity he seeks to reveal. Wittgenstein tells us and shows us that nothing in language use functions or has meaning in isolation, and he does not think he can explain how language works in a “coherent” book that will offer a codified theory, as he had earlier attempted. Language theory is “codification resistant,” as de Man says. And expert codification is too detailed to have practical utility for purposes other than specialist classification. This impracticality stems from, among other things, the copiousness and particularity of what Wittgenstein calls “language games.” Nongrammarians might be reminded of the effort of reading Halliday’s grammar, however useful it is for expert application. Halliday says at the end of his long career that “our interaction with our environment is so complex and multidimensional that there has to be a lot of ‘play,’ or indeterminacy, in the construal [of grammar] for it to be able to work at all” (Matter 63). This complexity reminds me of biochemistry, which presents an image of fantastic intricacy and complexity; the factors and forces effecting potential bonds between different elements of the periodic table are too complex and indeterminable to keep in mind for non-specialists. What de Man calls “codification resistance” lines up quite nicely with Halliday’s view on language theory’s indeterminacy (OL 266). It is not just the linguistic sign that is indeterminate, but also the grammatical mechanisms and the theories defining them. This thesis, too, may be charged with a degree of criss-crossing-in-every-direction as it  24  gathers together discourses that recommend a certain scepticism about linguistically mediated human understanding. I want to remind readers of the troubling epistemological implications and multiple opportunities for error resulting from analogical-cognitive operations that are mediated by a linguistic semiotic. My interest in a renewed linguistic consciousness coincides with the linguistic concerns of the evolutionary biologist Richard Lewontin, who worries about the misleading implications of the metaphors used by science in public discourses. But the linguistic-mediation problem is also a problem for science itself. Since language mediates even the terminology of hypotheses, linguistically entailed assumptions may occlude important evidence or produce false results. It is not just a question of empirical observation; language both construes and enacts, producing what we see. An increased consciousness of linguistic mediation and increased attention to language is warranted in all types of research. It is a question of considering the implications and assumptions underlying the actual terms in which research projects are imagined and the cognitive analogies on which selection of terms is based.34 1.5  the participants and other parts ~  Bacon appears in this dissertation’s cast of characters as exemplary of certain linguistic matters, not the least of which is his role as an intermediary between ancient and contemporary views. Nietzsche enters as the theorist who specifically links classical rhetorical theory to his own view of language’s metaphoricity. Derrida’s and Plato’s sceptical-rhetorical views of language and footloose stylistic have prompted this investigation as objects of study. The cursory stylistic analysis of their texts concluding this dissertation is the destination for its four preceding Parts. What I am calling the sceptical view of language (where sceptical means eyeswide-open) is elaborated through Voloshinov’s Marxism and the Philosophy of Language, with assistance from the findings of Sperber and Wilson’s model of communication and the surprising results of the research on reported speech. Linguistic-pragmatics scholarship, and 34 In a 2005 rhetoric of science collection, Fahnestock suggests that cognitive science make use of “the view of language implicit in rhetorical stylistics” as the basis for its neuroscience research examining the functions of language activities in the brain. Her belief that the rhetorical tradition’s “insights into language” should inform contemporary neuroscience reflects interestingly on the twentieth century’s “philosophy of language” failure to recognition the insights of classical rhetorical theory (Cognitive 164-65).  25  Sperber and Wilson in particular, provide a lexicon and an understanding of linguistic communication that illuminate the Derridian and Platonic stylistics. Halliday appears periodically as a language expert; and Wittgenstein’s exquisite, particularized examples of often obscure, abstract features of language use warrant his continued reappearance. While the authority of these scholars speaks for itself, Aristotle’s role here is a bit trickier. He might appear as an authority on language’s rhetorical operations; but he also proves to be someone whose views of language don’t seem to square. The gap between the linguistic implications of his utterances (what he says) and their style (how he says it) seems to indicative of his scientific motives. He serves as an object lesson in this thesis’ discussion of style. Part 2 presents some of the factors and participants featured in this thesis. Bacon’s “plain style” continues to serve as a pretext, while some of Voloshinov’s views come into contact with Halliday’s grammar and Sperber and Wilson’s ostensive-inferential model of communication. Part 2 ventures out on a sightseeing tour into several abstract dimensions of linguistic mediation like the role of affect in cognition, the subjective nature of the human perspective, and linguistically mediated reasoning processes. De Man’s notion of resistance continues to stage some of the difficulties of writing about language theory as our journey takes us to such occult locations as Silvan Tomkins’ theory of affect, William James’ “Sentiment of Rationality,” the cognitive use of analogy, Van Dijk’s analysis of the cognitive functions of specification and generalization, ancient maxims about knowledge and recognition, and the totalizing implications of the term theory itself. The leading role for language in these high-level and abstract epistemological conflicts makes each component of the mix a relevant consideration of how language works or doesn’t work. While most of these matters stem from ancient philosophical ruminations and will be familiar, even obvious, they seem to be frequently overlooked when it comes to thinking about language. Part 2 gathers together insights affiliated with the problems of linguistically mediated human reason and the resulting implications about language. Part 3’s account of late-twentieth-century language discussions have Nietzsche’s  26  important earlier insights as their destination. Part 3 exploits the late-70s controversy resulting from Searle’s response to the publication of Derrida’s talk “Signature, Event, Context” as a way of bringing Derrida’s ideas about context and citation into contact with the scholarship of reported speech. We witness a meeting of minds regarding important aspects of language use between different, formerly antagonistic, disciplinary inheritances in philosophy. Some of Derrida’s ideas about language mesh with conclusions of scholars investigating reported speech in various language-theory disciplines and departments (i.e., sociolinguistics, psycholinguistics, anthropological linguistics, grammarians, linguistic-pragmatics, discourse analysis and literary theory). As we shall see, these finding are also consonant with Voloshinov’s theorizing and Sperber and Wilson’s theory of communication. We witness the role of the promiscuous linguistic sign in human interactions, involvement strategies, and affective interests in speech. An interesting network of relations appears between, i) the Tomkins, James and experimentalpsychology discussions of affect in Part 2, ii) Fahnestock’s reporting of recent neuroscience research, iii) and the Voloshinov-Halliday view about the relationship between the sign and the psyche. This interdisciplinary work reinforces the ancient thinking-in-language concept and sets the stage for Nietzsche’s critique of language’s tropological nature. In Part 4 Nietzsche’s metaphorical turn leads to questions of style and clarity. Hume’s famous simplification of his Treatise, documented in Culler and Lamb’s Just Being Difficult: Academic Writing in the Public Arena, illustrates the vicissitudes of “the plain style” in Enlightenment philosophy; and various circumstances are called to testify to the role that the style-of-science myth has played in analytic philosophy. Aristotle’s clarion call for clarity provides an object lesson on style. When the performativity of direct discourse that we experienced in Part 3 meets the clarity of Aristotle’s analysis (or what has always seemed to be analysis) questions about just what style is blur. Style itself is part of the performative dimension of language use and the style-performativity-genre bond both prompts different reader responses and is motivated by different writer interests. Style’s intimate relationship with genre, which is style’s domestic home, helps clarify the role that genre expectations play in how  27  language works. Part 4 closes with a discussion of genre in anticipation of the Platonic and Derridian texts. Part 5’s analysis of Plato and Derrida concludes this dissertation. This study of the intertwined linguistic theories and stylistics of the Platonic and Derridian texts illustrates the image of language outlined in the previous Parts. This rhetorical-sceptical view of language results in what I call Plato’s and Derrida’s conversational stylistics. Part 5 takes stock of the history of mis-readings of the Platonic text, gauges the interpretive challenge its genre poses, and then explores the way that each writer exploits genre expectations to produce their distinct, ironic-conversational stylistics. Glancing back at assumptions about “propositions” from the “philosophy of language” we consider the unique ways that themes and theses are woven into the Platonic and Derridian texts; we observe the implicitness that characterizes these texts and the reader involvement they elicit. After gauging the way things are said in Derrida and Plato, the way that Derrida re-performs Plato’s conversational style, and the way that their texts “read us,” as Halperin says, we head towards the dissertation’s concluding analysis of what Plato says about language, and the way that his astonishing rhetorical mastery configures his scepticalrhetorical views of language.  28  PART 2 2.1  adventures of signs in human communication ~  The translator’s introduction to Marxism and the Philosophy of Language records Ann Shukman’s observation that “Voloshinov’s extreme contextualism leads him to a semiotic theory that is primarily sociological, and to a theory of language that emphasizes process rather than system, function rather than essence” (viii). Upcoming sections reveal the way that Voloshinov’s interest in linguistic processes, like the reporting of speech, lines up with the focus on actual language use from the practical advice of rhetorical theory’s ancient-handbook tradition to contemporary discourse analysis. Concerning “function,” the trans-disciplinary anatomy of Volshinov’s theorizing – semiotic, sociological and psychological – offers a glimpse into the manifold functions of language comprehended in his approach. The social origins of linguistic-semiotic phenomena shape and problematize language’s epistemological functions: complications arising from linguistic mediation central to Parts 2 and 3 of this dissertation. Voloshinov claims that the sign’s social origins have profound psycho-ontological impacts. These matters are addressed in this section by Tomkins’ and James’ reflections on the ties between affect and language, as well as the discussions of logic, reasoning-by-analogy, and part-whole relations. It is the social origins of signs that give rise to the complexity and multifunctionality evident in the work of linguistic-pragmatics and discourse-analyst scholars presented in Parts 2 and 3, and in the overlapping and intertwined behaviour of classical rhetoric’s appeals that are discussed in Part 4.35 This thesis is especially attentive to the implications of Volshinov’s “extreme contextualism” for genre and style. I have already alluded to the way that Wittgenstein’s text states and performs the “contextualism” that Shukman refers to: the way that the intellectual atmosphere and conversations of the 1920s that resulted in the Tractatus Logico-Philosophicus becomes the context for the “remarks” of Philosophical Investigations. On the other hand, Plato’s fictional dialogues provide situated conversations – in physical and social settings – 35  i.e., rhetoric’s appeals to the ethos (‘character’ or ‘habit’) of both speaker and audience, to the pathos (‘experience’ or ‘suffering’) of human life, and the logos (‘account’ or ‘story’) like the cause and effect topoi, etc.  29  characters with distinct personalities, and discussion of particular topics. Each ingredient of the dialogue contextualizes every word uttered. Alternately, Derrida’s famous parasitic writings, like Wittgenstein’s Investigations, grow out of other texts, though not primarily his own. Rather than attempt to inscribe codification-resistant language theory on a blank page (that is, without context), Derrida contextualizes his writings in the palimpsest of others’.36 We note here that the concept of context is itself extremely amorphous. It has a different look in different instances: depending on the size of its playing field, the level of language use under scrutiny, and the style of contextualism. If we imagine the copiousness and ponderousness that might have resulted had Wittgenstein attempted to summarize all his particular “remarks”– to codify, organize, and index their interconnections into a coherent book – we can see why these thinkers might have fashioned specific habitats (or “places”) for contextualizing their linguistic ideas. Voloshinov too finds a stage for his linguistic theory by setting it in the ideological domain: a “place” or location which is figured as the anthropomorphized universe of the sign. Like Bakhtin, he imbues signs with life and charges ideology with the animated energies of this enlivened sign.37 It is important to observe that Voloshinov’s ideology imparts the original broad range of perspectives in Marx’s definition, and not the narrowly political implications twenty-first-century readers might expect. Since in The German Ideology the term ideology represents “the production of ideas, conceptions of consciousness,” all that “men say, imagine, conceive,” and includes such spheres as “politics, laws, morality, religion, metaphysics, etc.,” ideologies are clusters of affiliated conceptions (47). The specifically political implications anticipated from both the title, Marxism and the Philosophy of Language, and its leading role for “ideology” are strikingly absent from the text.38 Though there are infrequent references to base, superstructure, and class struggle and doubtless political implications could be (and have  36  Glendinning observes that Derrida (like Voloshinov) defends “a particularly radical form of contextualism” (318).  37  For instance: “The outer sign originates from this sea of inner signs and continues to abide there, since its life is a process of renewal as something to be understood, experienced, and assimilated, i.e., its life consists in its being engaged ever anew into the inner context” (33). 38 The translator’s selection of comments, from Jameson’s review in Style and Jakobson’s Letters and Notes, reflect the text’s singular linguistic-semiotic trajectory. Jameson says “quite simply one of the best general introductions to linguistic study as a whole” (“Translator’s Preface” viii).  30  been) inferred from this book, I suggest that Voloshinov’s exploitation of ideology might be conceptually and stylistically opportunistic.39 What looks like political diffidence might be a rhetorical strategy. Admittedly the title’s reference to Marxist Ideology could also have provided Voloshinov (or Bakhtin) cover from the political persecution that Strauss suggests has produced stylistic “indirection” throughout the ages. But however much such indirection might have motivated the writer, at the same time ideology conceptualizes and images a semiotic site par excellence. It is a conceit or image of linguistic phenomena illustrating the twisted journey and adventures of the sign, and at the same time serves as a conceptual pretext for mapping the interdependence of idea and sign: a relation that crucially illustrates the linguistic mediation of human conception that preoccupies Nietzsche. “Without signs,” Voloshinov says, “there is no ideology” (9); “consciousness itself can arise and become a viable fact only in the material embodiment of signs” (9); the “sign is not only a reflection, a shadow, of reality but is also itself a material segment of that very reality” (11). The material reality of ideology is linguistic; and its phenomenal appearance has the form of signifiers, audio and visual objects in the world. Most importantly, the sign sits in reciprocal relation to the idea; the sign conditions the concept that conditions the sign. Signs and concepts are the part of reality that constitute our understanding of it.40 Modernity’s failure to recognize this ancient insight can be indexed by North America’s hysterical reaction to Whorf’s evidence that words shape thought and the absurd accusation of linguistic determinism levelled against him. The Whorf panic imagines a terrifying imprisoning sign whose rigid-determining capacity manacles conceptual creativity. This reaction bears the imprint of a Cratylian conception: that is, the code-metaphor model, the conviction that names refer to a determined referent, a blindness to language change, and the tendency to confuse the phenomenality of the sign with the phenomenality of the thing named. Thought must be independent from language in a Cratylian universe; otherwise, cognitive operations would be restricted by a finite lexicon. But 39 40  Some of the political interpretations are referred to in the “Translator’s Preface.”  Halliday explains this two-sidedness of the sign when he says that language “has to create a universe of its own, a parallel reality that is, as it were, made of meaning; [and] it is this ‘textual’ potential that enables language to be at the same time both a part of human experience and a model of it” (OL 244).  31  for Voloshinov “the meaning of a word is determined entirely by its context” and “there are as many meanings of a word as there are contexts of its usage” (79). The sign’s thoroughly social existence and ancestry embroils it in subjective human affective life. In Voloshinov’s view, its social gestation fosters the sign’s vitality and mutability and its refracting and distorting aptitude. He says, quite typically, “a sign does not simply exist as part of a reality – it reflects and refracts another reality. Therefore, it may distort that reality or be true to it, or may perceive it from a special point of view, and so forth” (10). This promiscuous-reciprocal-referential relation of sign to reality is also thematic in Halliday, Nietzsche, Derrida, and, as Part 5 suggests, in Plato. It is famously explored as “lenses” in Kenneth Burke’s “Terministic Screens.”41 Wittgenstein uses the related metaphor “spectacles” when he troubles over the way that logic becomes a lens for the philosophy of language (§103). For Voloshinov signs and concepts are inter-illuminating and in constant flux: characteristics of the human-linguistic-semiotic sphere thematized in Bakhtin and Plato. Words – whose entire realities are absorbed in their functions as signs – are a sensitive, mutable, context-dependent medium unsuited for the signal-type transmission entailed in code-metaphor conceptions of language. 2.2  Relevance Theory ~  On the other hand, Sperber and Wilson address an audience accustomed to (a-contextual, referentially-determinate) code metaphor models: an audience which imagines communication as descriptive-information exchange, and focusses on the truth conditions of propositions and speaker intentions as criteria for determining communication success. Their work in contemporary pragmatics – linguistics’ disciplinary response to the work of philosophers like the early Wittgenstein, Austin and Grice – disrupts many of their audience’s Cratylian assumptions about language.42 By adapting and expanding Grice’s notion of inference and his 41 42  in Language as Symbolic Action.  I am not suggesting that the entire audience is committed to code-referential, truth-condition determinations, a fact evinced by the emergence of the field of pragmatics itself. Code-referential-truth conditionality has posed some prime questions about language, and Sperber and Wilson try to take up these questions, sometimes answering them, but more often showing that they are the wrong questions. See Grice’s Studies in the Way of Words for an indication of this lineage and the model of inference he works with.  32  maxim of “relevance,” their Relevance Theory43 model of ostensive-inferential communication undoes habitual code-model notions of communication as sending information from one mind to another, and shows instead that hearers derive “implicatures” from utterances by making “inferences.” Sperber and Wilson play multiple roles in this thesis. On the one hand, their model serves here to dismantle the code metaphor and the primacy of truth-evaluated reference in communication. On the other hand, their theory is useful for understanding the workings of broadly conceived poetic or rhetorical effects and thus helpful in producing implicatures from the Derridian and Platonic texts, as well as being theoretically instructive about how these writers may have conceived what they were doing with words. The ubiquity of implicitness in the ostensive-inferential model and the important role for hearer inference illuminate how language works in sceptical (eyes-wide-open) language theory. I focus crudely on several salient elements in relevance theory’s ostensive-inferential model of communication useful for a very basic understanding of it. Sperber and Wilson demonstrate that: 1) linguistic communication is an interpretive-inferential phenomenon, and 2) some relative of logic (which is a red herring in the code account) plays a limited though crucial role in producing propositional implicatures used in the interpretive-inferential processes; 3) referentiality (another red herring) plays a less significant role than say, the importance of hearers’ participation in what are often vague communicative efforts. In other words, reference is clearly subordinate to the importance of hearers making their own inferences. And 4) communicators are little interested in being informed by means of explicit reference.44 Actual language use is all about implicitness. Rather than communication being the simple transfer of a message from one mind to another (a passive recipient), hearers are active participants. The idea of informing someone of a message has shifted ground and metamorphosed into hinting at the sort of thing a speaker and hearer might want to discuss.  43 All references to Sperber and Wilson in this dissertation are to their Relevance: Communication and Cognition except as noted. 44 I use the term explicit in its common sense meaning here, and elsewhere until Part 5, when it becomes necessary to acknowledge that the concept of verbal explicitness entails determinate reference and has the same fraudulent status as verbatim explored in Part 3.  33  In the ostensive-inferential model of communication, meaning-making is a mutual affair, with speakers who point at something and hearers who first determine what is being pointed at and then interpret what about it the speaker finds interesting enough to point out to them. The ostensive-inferential model includes a kind of deductive (i.e. logical) function performed by the hearer (along with other heuristics) in the inference process. Utterances are enriched or expanded to propositional form (by inference) according to intuitive logical principles (and other heuristics) based on the context; and what the hearer already knows (a contingent factor) are premises in the enriching inference. Each inferentially enriched utterance, in this model, serves as a bit of the material that hearers must use in subsequent inferential activities – activities which are highly interpretive despite deductive workings. This inferential process culminates with the most relevant implicature (an estimation not a deduction) that the hearer can derive. Not only are speakers “presumed to aim at optimal relevance not at literal truth,” but the guesswork involved in enriching utterances to propositions and the range of things that the hearer already knows makes the inferential activities in this model relatively indeterminate from the outset (Relevance 233). Relevance drives truth-conditions and reference to the sidelines. This fuzzy process of inferential guesswork, repeatedly performed at several levels of the communicative process (by active hearers/readers), has salience in the ostensive-inferential model of communication. For Sperber and Wilson “human intentional communication is never a mere matter of coding and decoding” (174). This is in part because human “languages do not encode the kind of information that humans are interested in communicating”; instead language users are interested in what languages do, and what they can do with them (174). This observation about what we are interested in communicating reminds us that communication processes are shaped by their primarily social motivation: human interests in interacting and participating. Voloshinov too emphasizes the social conditioning of signs and linguistic expression. He argues that human linguistic signs and the speech genres that stylistically shape utterances have completely social origins (20). The inferential model of communication discovers that speakers  34  are not informing (passive) hearers of a message content, but pointing to or hinting at the kinds of things they are interested in discussing and drawing attention to certain things so that the (active) hearers can make inferences. Sperber and Wilson describe this process as “making assumptions manifest” and bringing assumptions into the “cognitive environment.”45 So, for instance, even when an explicit interest in closing the window motivates communication, the question, “is it cold in here?”, might be a preferred method of expressing the interest. And the speaker’s preference for the question itself can trigger a chain of implicatures. Such preferences are a fortiori the case when the communicative intentions are less certain, as Sperber and Wilson suggest they frequently are; they say that the effects of most forms of human communication are “vague” (55). When other factors like being polite, just having something to say, or just wanting to respond are at stake, communicative aims may be much more general, not nearly as determined as “closing the window.” Implicatures are not explicitly stated by speakers but inferred by hearers (or readers) based on what is said (and not said). Hearers (or readers) attempt to determine how the implicatures of an utterance achieve relevance: they gauge the relevance of implicatures – not just what is said, but also the implications of its being said, of how it is said, and its being said in this particular context by this particular person. In Sperber and Wilson’s terminology, hearers seek optimally relevant implications from the assumptions that have been made manifest. The role of the hearers in making inferences from the assumptions that have come into the cognitive environment (including what the hearer already knows) is paramount in this social and interpretive activity of communication. In fact, Sperber and Wilson find that people don’t enjoy being told things that they already know: A speaker aiming at optimal relevance will leave implicit everything her hearer can be trusted to supply with less effort than would be needed to process an explicit prompt. The more information she leaves implicit, the greater the degree of mutual understanding she makes it manifest that she takes to exist between her and hearer. Of course, if she overestimates this degree of mutual understanding, there is a risk of making her utterance harder or even impossible to understand. It 45  Note, Sperber and Wilson say that “assumptions are unlike knowledge in that they need not be true” (40).  35  is not always easy to strike the correct balance: even a slight mismatch between speaker’s estimate and hearer’s abilities may make what was merely intended to be helpful seem patronising or positively offensive to the hearer. (218) It is better to repair a misunderstanding than to over-inform: so much for code metaphors of message delivery, sociality is the organizing principle here. The speaker hints and the hearer guesses, an interaction that looks a lot like a “game.” One of Halliday’s pillars is that language is a social semiotic: that it always has both interpersonal and ideational/experiential meaning. As we shall see in Part 3, Tannen has collated research demonstrating that “involvement” and “solidarity” – entirely social-affective motives – are what is at stake in conversational discourse. Social (affective) motives, according to Voloshinov, shape all semiotic components and structures. Finally I wish to draw attention to the distinction that Sperber and Wilson observe between a communicative intention and an informative intention (which, however problematic, is nonetheless useful). Communicative intentions, which pertain to a social sphere, can be made manifest ostensively: that is, by non-linguistic demonstration or showing, like pointing, or looking at something and making a face. Sperber and Wilson conclude that human communication is conducted in this ostensive-inferential sphere; communication pertains to showing or demonstrating, making assumptions manifest so that a hearer can look and try to determine what implicature is most relevant in the context (49). This speaker-hearer activity squares precisely with the findings of research on reported speech, as we shall see. It is here that we witness the final collapse of the code metaphor because the utterance may signify something (like “it’s cold”) entirely other than its referent (“close the window”). The distinction between the referentially-descriptive and the ostensive-inferential is parallel to the distinction between the constative and the performative in Austin. Since language often performs multiple functions (at the same time), frequently with divergent aims, and since it is often used not descriptively but as a pointing device (or a demonstration, as we shall see with reported speech), it cannot be taken at face value (or read as description), as the style of non-style insists on doing. Language use in communication has more to do with pointing, inference and mutual  36  interest than message transfer. In this context, Sperber and Wilson determine that language is needed for the cognitive function of information processing rather than communicative functions strictly speaking (173). “Languages,” they say, “are indispensable not for communication, but for information processing; this is their essential function” (172). Information and what it might be are other problems that we don’t have time to explore here. Information processing clearly is employed in the inferential processes of linguistic communication (e.g., she is asking is it cold in here and looking at the open window, perhaps she wants me to close the window, or maybe she’s worried about the dust coming in). If we imagine “information processing” as a linguistically dependent function using a socially conditioned semiotic material we end up with an account that resonates with the semiotic milieu of both Halliday’s “construal” of experience (OL 276) and Voloshinov’s “consciousness” and “psyche” (9, 26). Sperber and Wilson’s findings suggest that it is not specifically (social) communicative impulses that depend on specifically linguistic signs, but thinking itself that requires them. Linguistic communication is a complex interface of social semiotic and affective human motivations that we explore shortly through the ideas of Tomkins and James. The assumptions that flow from unexamined conceptions of communication have doubtless bedevilled language theory. Ellis considers “the assumption that the purpose of language is communication” to have been a major misstep in twentieth-century language theory (15). He associates conceptions of communication with code metaphors, pointing out that Whorf observed the error of “supposing the function of language to be only the COMMUNICATION of thought” (20). Halliday adds that both Firth and Hjelmslev “rejected the notion of language as a system of communication; they saw language as constructing reality not passively reflecting it; and they distrusted the pseudo-science and the spurious rigour of models from formal logic (‘rigor mortis’ was Firth’s way of characterizing it)” (OL 242). Halliday says that his mentor, Firth, “explicitly rejected the definition of language as a means of communication and was highly critical of those who imported into linguistics a conception of language derived from communication theory” (OL 242-43). Sperber and Wilson have remade  37  communication theory in a way that matches the insights of linguists like Whorf and Halliday. Of course, none of these scholars are disputing language’s use in linguistic communication. But communication is not all language is used for. It is used for thinking, identification and understanding the world – epistemological purposes. And it is unexamined notions of communication, intuitions leading to code and conduit conceptions that are so confounding to language theory. Nevertheless it is, according to Voloshinov, language’s sociality – its social conditioning and social origins – that shapes linguistic-semiotic phenomena and governs the operations of the entire linguistic sphere. This means that epistemic operations employ socially conditioned, socially fabricated, and socially saturated semiotic materials. Perhaps the disjunction between language’s social origins and its epistemological-thinkinginformation-processing functions is one of the sources of its complexity and codification resistance. Code does not survive Sperber and Wilson’s re-figuring of communication (though the boomerang tendencies Ellis observes cannot be overstated). Reference survives by adapting to indeterminacy; and description moves over to make room for other types of linguistic operation. The epistemological uses of language is of particular interest, as it relates to Bacon and the vicissitudes of the style of science. However consistent or inconsistent Bacon’s views of language may or may not have been, they were eventually displaced by superficial code-conduit models of transparency. To the extent that Bacon’s rich understanding of language was reframed in this simplistic manner, the seventeenth century’s exuberance about natural history, the freezing point of mercury, and the transit of Venus seem to have drown out Bacon’s linguistic insights. This cursory consideration of Bacon raises questions about what kind of “advancement” regarding the linguistic paradigm took place during the Enlightenment and whether a Kuhnian backsliding might not have occurred in linguistic consciousness. Voloshinov complains about the “mechanistic conception[s] of linguistic necessity” that had become standard in the (“mainstream”) linguistics of the early 1920s (81). Did they get their start in the seventeenth century’s knowledge quest when the aspirations to secure knowledge of the world  38  and to theorize about language seemed to veer in opposite directions? Perhaps anxiety about language’s epistemological frailty prompted the Royal Society’s resistance to Bacon’s linguistic theories and millennia-old rhetorical insights in favour of the excitements of new knowledge. 2.3  historical context: Bacon and the “plain style”~  Zappen’s work on Bacon’s rhetorical theory documents an affiliation that history and rhetoric of science scholars (circa 1920-1950) make between Bacon and the “plain style” of “positivistic science” (Historiography 75). The positivists simply seem to lay claim to Bacon as the father of science and ideologically reify his purported plain style as the style-of-non-style. But the twentieth-century story changes after the positivists. Zappen observes that as early as the 1960s this view was rejected by historians of science and scientific rhetoric, and replaced by an understanding of the complexity of Bacon’s theory of style. These later historians observe that Bacon’s theory “is actually complex, that it embraces several methods of presentation and several styles, not just the plain style” (Historiography 76-77). However, it is not clear that Bacon advocated such a style at all, let alone practised it. Zappen’s reference to “the plain style” seems to stem from inferences based on Bacon’s critique of ornament and expressed admiration for “plainness” rather than a prescription advocating any particular style. Interestingly, until Lipson, the question of what the “plain style” might actually be does not seem to arise in these history and rhetoric of sciences discourses. The prose practices and intellectual climate of the early Renaissance together with the contemporary teaching of writing or rhetoric would provide a context for understanding what prompts Thomas Sprat, in the History of the Royal Society, to conclude “that eloquence ought to be banished out of all civil societies as a thing fatal to peace and good manners” (111). The competing linguistic discussions of the period that Kenneth Graham catalogues suggest that there were considerable contemporary debates and disagreements about trends in manners of discourse.46 Sprat’s prognostication on the ill effects of ornaments – “the myths and uncertainties” that “specious tropes and figures have brought on our knowledge” – contribute to 46  See Graham’s The Performance of Conviction and Stanley Fish’s “Epilogue: The Plain Style Question” for indications of the way that these trends have been analyzed.  39  a substantial contemporary debate (112). Whatever strains of rhetoric, whatever religious and scientific movements, and whatever overlapping and divergent views (spawned in this burgeoning print-literacy era) that occasioned Sprat’s reaction against “eloquence” and “ornament,”47 it is clear that while Bacon sometimes expresses concern about undue ornamentation, he also explicitly approves of and uses rhetorical figures in his text, as Lipson observes. Somewhere along the road the determinate-reference-code model is projected onto notions of “plainness.” Although Bacon (infrequently but admiringly) employs the term plain when talking about speech, it is usually in the context of hoping that his own efforts will be “set forth plainly and perspicuously.” But it sometimes appears in prescriptions that seem more concerned with “honesty” and “openness” than “simplicity” or “clarity.” For example, “plainly” seems to mean “openly and “honestly” when Bacon advises that “if in any statement there be anything doubtful or questionable, I would by no means have it suppressed or passed in silence, but plainly and perspicuously set down by way of note or admonition” (IX Preparative Towards a Natural and Experimental History). Lipson concludes that “confusion over whether Francis Bacon was the father of plain prose probably stems from the different possible meanings of plain style” (144). Graham finds that there are competing views of “plainness” in the Renaissance: 1) an antirhetorical plainness associated with “a certain claim to truth” characterized by an “insistent demand for certainty,” and 2) a rhetorical plainness which seems to constitute a call for a common (or democratic) language and entails a sceptical-rhetorical and socially negotiated understanding of truth (14-17).48 While Bacon’s views on language seem to fit him in the latter class, as Lipson indicates, 47  While it is not clear what definition of “ornament” might have been circulating – and there were likely as many different ideas of what ornament was as there were persons thinking about it – even Sprat admits that ornaments “were at first, no doubt, an admirable instrument in the hands of wise men” (112). Fahnestock equates “ornamentation” and “forcefulness”; she states that “[early modern] rhetorical manuals placed the figures of speech under this heading, for the figures were seen as sources of distinction or ‘ornatus,’ that vexed term usually translated as ‘ornament,’ rather than, as Vickers and others have argued, something closer to “armament [ornamentum], the weapons and gear that make a warrior or speaker effective” (Stylistics 218). 48  In an ironic oversimplifying gesture, Fish argues that the distinction distinguishing the wide variety of the era’s intellectual trends is an epistemological one: those who, as he puts it, “see through a glass darkly” and those for whom the “glass has been made clear by reason” (Epilogue 381). The distinction is framed as “an assumption of the mind.” He places Bacon on both sides of the assumptions-of-the-mind divide apparently on the basis of comments made in Advancement.  40  what is germane here are the varying meanings of plain. The meaning of plain is, according to Graham, “as plain as the meaning of ‘true,’ which is in fact its closest synonym in the sixteenth and seventeenth centuries” (1). The valence of “true” in “plain speech” helps us understand, as Graham observes, how “even such Shakespearean wordsmiths as Hotspur and Othello appear to believe that their language is plain” (2). This dimension of the term “plain” helps make sense of the “proliferation of [Renaissance ] plain styles” which are “catalogued by modern critics” and which are, arguably, not very plain to a twenty-first-century ear: “native plain style, classical plain style, Puritan plain style, Anglican plain style, Bacon’s plain style, Herbert’s plain style, Restoration plain style” (Performance 2). “Every claim to truth,” according to Graham, “is likely to call itself plain at some point” (2). Both Fish and Graham also highlight the “making evident” valence of “plainness” which Graham calls “a texture of obviousness” that he affiliates with repeated claims about scriptural “meanyng” (12). The career of the scientific plain style might begin here in its prior affiliation with truth and the discovery of (purportedly obvious) truth. However, truth is not self-evident, neither simple, clear nor obvious: and neither is the meaning of plain style. While it is not clear what the “plain style” is (or was), it has become affiliated with conceptions of determinate reference and description, and possibly a style that pretends not to perform. In Part 3 we will see the way that reported speech is paradigmatic for language’s performative dimensions, what Halliday calls its “semogenic” powers that combine “construing” and “interpersonal” functions. While Bacon’s empirical interests may have motivated his aspirations to speak “plainly and perspicuously,”what is interesting is that his many comments on language which have important epistemological consequences do not survive his purported prescriptions.49 Instead, his remarks about openness and his aspirations for the success of his own writing have morphed into a mythical style of science involving simplicity, clarity and transparency. But Bacon  49 Graham claims that “in Neostoic doctrine, right reason replaces Scripture as the principle source of plain truth,” hence “the reformed claims about the authority of Scripture parallel the claims made on behalf of the authority of reason in Neostoic doctrine” (6,8). The anti-rhetorical reform and Neostoic claim to truth and call to plain speech are thus distinct from the humanist rhetorical tradition’s dialogical plainness. The humanist call for plainness was a call for common language as a mechanism of communication: if Bacon fits in this category, it reflects interestingly on his conception of empirical science.  41  resolutely presents the epistemological problem of linguistic mediation: The syllogism consists of propositions – propositions of words; and words are the tokens and signs of notions. Now if the very notions of the mind (which are as the soul of words and the basis of the whole structure) be improperly and overhastily abstracted from facts, vague, not sufficiently definite, faulty – in short, in many ways, the whole edifice tumbles. (The Great Instauration) Here the definition problematic (linguistic recursion mise en abîme) together with the interplay between ideas or conceptions and words radically undermines syllogism (a frequent brunt of Bacon’s derision). Isn’t it interesting that the linguistic implications of this rejection of syllogism – the implications that referential failure is probable and our understanding faulty – were overlooked? Fish’s “The Plain Style Question” makes clear that the diversity of categories of styles prevailing in the seventeenth century have confounded repeated classification efforts, and that there is no agreement whether science or religion played the biggest role in stylistic prescriptions. In a light-hearted ironic tone, Fish offers yet another pair of seventeenth-century styles – this time in terms of reading experiences: “the experience of a prose that leads the auditor or reader step-by-step, in a logical and orderly manner, to a point of certainty and clarity” and “the experience of a prose that undermines certainty and moves away from clarity, complicating what had at first seemed perfectly simple, raising more problems than it solves” (378). (We will return to this affiliation of certainty and clarity.) Overall Fish observes that, as a prescription for writing, officially “the plain style wins the day” over a prose carefully qualifying its own certainty (379). Notably, since this prescription (as Graham’s catalogue indicates) means nothing in particular and so many possible things, the “plain style” becomes an empty sign. Perhaps this is how the scientific plain style becomes mythologized. The naivete with which the idea of a plain style has been received seems to go together with code-model intuitions. We can see that it might have been here that the materiality of the signifier started to become transparent, and the sign become confused with the materiality of what it signified – a Cratylian dream, where the sign is essentially alloyed to its unique referent.  42 2.4  epistemological quagmire: affective rationality ~  Having gauged the sign’s promiscuity (its freedom from referential restraint) and the Enlightenment-bred resistance to this insight (despite Bacon’s contemplations about the referential instability of words), we turn now to other epistemological implications of linguistic mediation, and consider the influence of commonplace assumptions about language’s rationality. We begin by contemplating language’s role in cognitive functions and the influence of the sign’s affective sociality on supposedly “rational,” “logical” mechanisms. Bacon’s above noted concern that potentially “faulty” “notions of the mind” are the “soul of words” places him in a long line of language sceptics aware of the worrisome epistemological implications of linguistic mediation (from Parmenides to Derrida). Fahnestock explores this word-mind intimacy in her work on topoi (that we discuss shortly) and in her work on recent neuroscience findings (discussed in Part 3); this intimacy is also apparent in Voloshinov’s endorsement of the ancient insight that we think in language (or at least some kind of semiotic).50 He observes that “the reality of the inner psyche is the same reality as that of the sign. Outside the material of signs there is no psyche; there are physiological processes, processes in the nervous system but no subjective psyche” (26). Interestingly, psychologist Silvan Tomkins will even link some of these physiological, nervous-system processes with the psyche and language. The biological responses, that Tomkins calls “affect,” and that constitute “the motivation system,” reveal an interplay between language, affect and thinking, intolerant of code-metaphor models and crucial to understanding how language actually works. It is relevant to this dissertation that recent discussion of this psyche-language union also recalls classical rhetoric’s appeals – socially-inscribed affective responses that speakers tap into for producing persuasive effects. These include, for instance, appeals to ethos which in order to persuade count on establishing the esteem of the speaker or identification between speaker and  50  Parmenides’ epistemological quagmire concerns “deceitful” words which arbitrarily divide what is one into many. He says, “thinking and the thought that it is are the same; for you will not find thought apart from what is, in relation to which it is uttered” (8.34-35, Robinson 110). Gallop remarks that “‘speech’ and ‘thought’ are coupled continually throughout” Parmenides’ text (“Introduction” 31n). This language-thought union which I believe was commonplace in the intellectual climate of ancient Greece appears in Plato a century later: for example (see Theat 190a, Soph 263e, Phaed 266b, see also the proto-Lacanian account in Phil 38e-39a).  43  audience. Jennifer Richards’ Routledge handbook Rhetoric identifies affect as the defining feature of rhetoric: “rhetoric is concerned with the affective power of language and with describing and classifying the devices that produce emotion, or develop a logical proof, and so sway the judgement of an audience” (121). It should not be surprising then that Voloshinov’s linguistic theory includes a chapter titled “The Philosophy of Language and Objective Psychology.” His purpose is not to analyze the individual psyche, but rather to trace the life cycle of the (social) sign – the way that it is shaped by its travels within an individual psyche, out into the world, and back in again – the socio-psycho gestation of the sign. This journey charts the sign’s subjective and inter-subjective social conditioning and accretions, providing a companion account to the absence of centre, the adventures of the trace and linguistic supplementarity that Derrida treats. For Voloshinov the individual “consciousness” is dependant on the social semiotic medium of language. In his account, the role of the human psyche and its social network in shaping the sign also indexes the psyche’s use of the sign. This reciprocal language-psyche connection signals the relation of language to an affective realm of subjective and intersubjective desires, interests, feelings and beliefs – a language-psyche-affect conjunction. Halliday’s views fit perfectly with Voloshinov’s here, as they do with Nietzsche’s. For Halliday “language structures all of our experience, it enacts all of our interpersonal processes” (OL 251). “Language is the essential condition of knowing, the process by which experience becomes knowledge” he claims; and “we might interpret learning as something that is inherently a semiotic process” (Learning 94). It is worth taking time to absorb the radical claims that Halliday is making about the linguistic nature of consciousness, as well as the way this linguistic transformation of experience into meaning not only construes or constructs a referential reality, but also acts in it. Language is simultaneously a mode of acting and a mode of thinking: Language does not ‘reflect’ our thought processes; it actively creates (‘construes’) the categories and relations with which we think. Here ‘transforming matter into information’ means transforming our experience into  44  meaning: both what we experience as ‘out there,’ in the world that lies round about us, and what we experience as ‘in here,’ as going on inside our own bodies and our own consciousness. Hence a theory of grammar (a ‘grammatics’) is a theory of how language does this. If that was all that the grammar achieved, we might say that it was just another system of information. But this is only one part of what the grammar does. As well as construing experience, the grammar also enacts our interpersonal relationships; and here it is not a mode of thinking – it is a mode of acting. When human beings evolved the resource of language, it served from the beginning not only as our means of information (the ideational function) but also as our means of interaction (the interpersonal function); and the two are inseparably bonded – you cannot have one without the other. Every act of meaning is both construal and enactment at one and the same time. (OL 276) Halliday here helps us see the potential choreography between the inter-subjective desires of the psyche and the social interpersonal functions of language that roll out at the same time as the construal of what we might imagine as an objective-referential reality that is out there, that we might think we are describing. Halliday thus highlights the thinking element already implicit in Voloshinov’s language-psyche-affect bond, which, as noted above, has a corollary in psychologist Silvan Tomkins’ script theory. Tomkins’ seminal studies of affect, which result in his theory of the affect (or motivation) system, have roots in Darwin’s pioneering work in The Expression of the Emotions in Man and Animals. Tomkins’ complex, hard-wired, biological-feedback-system model of affective responses both modifies and displaces Freudian drive theory. It foregrounds the discrete role played by the physiological-psychical, nervous-system responses he calls “affects” and their crucial contribution to human motivation. Briefly put, human affects are physiological responses with a psychic impact like terror, anger, shame, distress, startle, joy, and excitement. They are the responses that motivate human behaviour.51 The objects and aims of affects, unlike those of drives, are mobile and flexible; they respond to contingent factors impinging on them, and are governed by reciprocity. An affect can change, displace and replace its objects, attach to a new object, and join with other affects to act in concert. Tomkins describes them as multiform 51 McIwain categorizes Tomkins’ affects more specifically as six basic affects (interest-excitement, enjoyment-joy, surprise-startle, fear-terror, distress-anguish and anger-rage), one affect-auxiliary (shame) and two drive-auxiliaries (disgust and dissmell) (500).  45  and multifunctional because they can play different roles depending on incidental interactions and on whether they join forces with other affects and/or drives. The stellar role of affect as the primary source of human motivation and social development gives an added eminence in human cognitive processes to what has been considered passion or emotion – non-rational spheres. It creates a new picture of an affectively conditioned human rationality. In Tomkins the cognition and motivation systems are intertwined providing a novel image of what human reason is that is consistent with recent psychology studies (examining, for example, economic assumptions about rational decision-making).52 Tomkins’ studies on the motivation system lead to the discovery of a relation between linguistically formulated “scripts” or stories and human affective responses to situations and events. A script is a linguistically (or semiotically) remembered account in which affectively laden scenes undergo psychological magnification (a semiotic-cognitive interaction) connecting them to other affectively laden scenes.53 A script then becomes “a set of rules for predicting, interpreting, responding to and controlling a magnified set of scenes” or a series of events (Shame 179-80). A script constitutes cause-and-effect relationships between scenes or situations, which determine how an individual will respond to and interpret a situation.54 As predictors and determiners of human affective responses, Tomkins’ linguistic scripts, according to Ed de St. Aubin, anticipate the recent and much celebrated “emergence of narrative  52  I am thinking in particular of Daniel Kahneman’s Nobel prize in economics for examining the foundationaleconomics assumption that human behaviour is governed by self-interest and that humans are capable of “rational” decision-making. Kahneman’s findings about the non-rational basis of human, economic decision-making have since been repeatedly corroborated and become commonplace in public discourses, particularly in the aftermath of the 2008 stock market collapse. 53  Components of Tomkins’ theorizing have become current (e.g., script theory, features of affect and its discreet motivational role) whether or not his authorship is recognized. Researchers in psychology and psychiatry familiar with Tomkins’ work observe that he has not yet had the full impact on current scholarship that he is likely to have in the future (de St. Aubin, McIlwain). Discourses on the relationships between linguistically formulated ideas, affect and cognition are commonplace in contemporary, psychiatry, psychology and sociology. But as McIlwain notes Tomkins is under-acknowledged, many don’t cite him where one would expect and few have read him in the original (500-01). In the context of my work it is ironic that his “dense and difficult to follow” writing is partly blamed for this uptake delay (de St. Aubin 1). 54  Like any linguistically formulated understanding memory scripts can take place in varying relations to consciousness. Some may be unconscious like Tomkins’ example of what he calls an anti-toxic script: “a child who suffers massive intimidation at the hands of an alcoholic father who often beats him may retreat to a defensive introversion, seeking to hide from and to avoid or escape dreaded violence. In an antitoxic script the child could become very withdrawn to the point of catatonia” (190). An “addictive” script which transforms “a sedative into an end in itself,” on the other hand, might be quite conscious: “the cigarette addict exaggerates the necessity of smoking a cigarette for his well-being as the miser inflates the necessity of money” (192).  46  psychology”55 (1). Script theory thus links affective (biological) responses to linguistic formulations (accounts or stories) operative in the psyche. The reciprocal self-validating and sometimes self-fulfilling role that these linguistic scripts play between the motivation system and the psyche seems to corroborate Voloshinov’s conviction about the sign/psyche relation. One might, for example, experience an instantaneous, semi-conscious perception of an affective response (like excitement, shame, or distress) that triggers a kind of movie-like-memory that runs something like “this always happens when . . . .” This would be the surfacing of a Tomkins-script, a set of scenes which have been carefully edited and linked by aggregate affective pressures, and then stored to direct future interpretations of experience. The interplay between a person’s affective response and their (linguistic) interpretation of that situation reveals the way in which feelings or sentiments permeate what are thought of as rational cognitive functions. We might think that an “account” (linguistically formulated interpretation of a cause-and-effect relation) is rational, but if it is affectively directed and produced, we have to ask what “rational” means. What criteria distinguish rational from non-rational interpretations of experience? Not only does the abstract concept of “rationality” become amorphous, but what we think of as rational – cause and effect relations, patterns of thought – also acquire a nebulous, logical status. Like any linguistic account scripts are subject to constant revision; if linguistic scripts determine what is rational, then rationality too is constantly revised. Rationality is the signature quality of human reason. According to the Oxford Companion to Philosophy reason is the faculty of truth-seeking or problem-solving, and logic is sound reasoning (Honderich, ed.). However, Tomkins’ script theory suggests that the rational and logical live in a never-imagined, close proximity to the affective realm. The rational and logical are always motivated by human affective responses. Tomkins complicates naive  55 Narrative psychology research into the way stories shape lives “the way human beings deal with experience by constructing stories and listening to the stories of others” challenges “the notion that human activity and experience are filled with ‘meaning’ and that stories, rather than logical arguments or lawful formulations, are the vehicle by which that meaning is communicated. . . . Sarbin (1986) proposes that ‘narrative’ becomes a root metaphor for psychology to replace the mechanistic and organic metaphors which shaped so much theory and research in the discipline over the past century” (Hevern).  47  idealizations of rationality or reason, as such. In the “Introduction” to Cognition, he identifies the problematic role that the idealization of “reason” as a “divine spark” plays in the landscape of psychology (2). For Tomkins everything is biological and evolutionary. He believes that there are a very wide range of types of knowing – from the genetic and neuronal to complex human physical and mental problem solving. But “knowing” is widely varied even within the “restricted case of thinking that is oriented toward problem solving, there are critical differences between solving mathematical problems and learning how to hit a golf ball effectively” (6). Tomkins’ cognitive system is firstly intertwined with the motivational system (apparently he refers to the cognitive system as a discrete system provisionally for the purposes of analyses). The cognitive and the motivational systems sit in complex relations to each other: related by partial dependence, independence and interdependence that varies according to contingent factors and “the specific state of the whole system at any one moment” (7). Rather than some generalized and unexamined conception of rationality or reason (as some kind of abstract-logical deductive ability), Tomkins life-long empirical studies of human motivation in psychology lead to a conception of cognition as a kind of networking function fed by a variety of specific biological mechanisms that is interdependent with affective mechanisms. Cognition is not just operational in abstract cogitations about human ethics or how language works, but also in the complex motor-perception-memory mechanisms required to hit the golf ball and in the operations of the nervous system.56 For Tomkins cognition is a kind of organizing, connecting, information processing that is built into various biological mechanisms and always intertwined with the motivational (affective) system. Against unexamined ethereal conceptions of rationality or reason dominating common-sense understandings of human behaviour and language use, Tomkins’ affective responses activate and shape the “interpretation” that produces an account (whether that “account” is the conclusion of a scientific experiment, an estimation of how far you have hit the ball, or the explanation for your 56  He says, for example, that “inasmuch as the neuron is the most elementary organized unit of the information processing mechanisms, it serves well to illustrate some of the general features of cognition, as well as the degree to which the simplest parts of a cognitive system share in the characteristics of the more complex cognitive mechanisms, including those of the total cognitive system” (Cognition 34).  48  backache). The role of language in storing the interpretation of affectively shaped accounts in Tomkins’ theory, together with the increasing recognition of Tomkins-like views in psychology, corroborates Voloshinov’s views about the psyche being a semiotic domain (the zone of contact between the sign and the psyche’s emotions).57 The emergence of linguistic scripts from “the totality” of cognitive and affective mechanisms presents an image of cognition that is a complex of interdependent networks. These frequently involuntary, automatic mechanisms undermine common-sense assumptions about what “rational” means (Cognition 9). Motivation’s affective nature lingers ubiquitously in our interpretation of experience, complicating rational accounts or explanations and communication’s fuzzy inferential operations. De Man’s metaphor of resistance is helpful because it signals repression, something recognized as a “non-rational” element in human cognitive operations. The temptation to assume that especially scholarly judgement is reasonable, “logical,” and consistent tends to overlook the role of affect in cognition. On one hand, at a high level of generality, this assumption overlooks the way that the attachment to ideas is an affective response. It overlooks the way ideological attachments – like faith in the unmediated access to empirical knowledge – influence how we interpret things, what we see to start with, and what we think language is. On the other hand, at a microscopic level, the assumption that judgements are rational and logical overlooks the way that what we consider rational (a cause-and-effect script) is affectively shaped. If all linguistic presentations are susceptible to affect, the domain of the rational needs to be re-thought. We need to make room for the salient role of motivation in rationality. This re-thinking affects a whole series of interrelated assumptions about thought, language, and logic: that human thought is rational (without consideration of linguistic mediation and the role of motivation), that language is a transparent and impartial epistemological medium (without 57  For the currency of this connection between language, thought and cognition see a 2004 article on inner speech by Akira Miyake, et al. They write “language plays an important role in everyday living. Although not the only medium, the central role of language in human communication is incontrovertible. The utility of language is not restricted to interpersonal communication, however. It can also shape and support intrapersonal thinking processes in various ways (Carruthers, 2002; Gentner & Goldin-Meadow, 2003; Sokolov, 1972; Vygotsky, 1934/1986). One such intrapersonal cognitive function attributed to language is the regulation of thought and action (Barkley, 1997; Luria, 1961). In particular, at least among adults, inner (or subvocal) speech is considered to play some role in regulating one’s behavior (Carlson, 1997; Sokolov, 1972; see also Gruber & Goschke, 2004). In this sense, inner speech (and, more generally, language) can be viewed as a contributing factor to efficient executive control” (124).  49  consideration of its suitability for “accurate” linguistic identification and accounts), that language functions according to a coherent, logical, rule-governed grammar (without consideration of its socially conditioned, affectively motivated origins). De Man explores the affiliation of the ideas of (common-sense notions of rule-governed) grammar and logic in “Resistance to Theory.” He notes that there has been a “persistent symbiosis between grammar and logic,” that “the grammatical and the logical function of language are co-extensive,” and in fact “grammar is an isotope of logic” (14).58 Grammar, from this perspective, would be a type of logic – a confusion of precise mathematical operations with linguistic ones. The affectively motivated logics of linguistic scripts taint idealized images of a logical linguistic grammar. In fact it might be observed that the affectively laden linguistic sign makes a game out of “logical” grammar – an approach to language taken up by both Wittgenstein and Derrida. The role of logical operations in human language is usefully qualified and delimited in Sperber and Wilson’s theory as we shall shortly see. As well, we will consult Halliday’s views of what grammar actually is, in order to better assess this rule-governedgrammar-logic affiliation. What is at issue here, is the imprudent severing of affect and cognition (and of passion and reason) in our conception of rational thought. This divorce disrupts apparently complex interdependent networks of “logical” and affective functions (that lack simple binary polarity). We might do well to recall that human reason is never imagined as devoid of passion in Plato. Recall, also, that Aristotle’s rhetorical appeals are to ethos, pathos and logos; about logos he says “persuasion occurs through the arguments (logoi) when we show the truth or the apparent truth” (On Rhetoric 1.2.6). Logos persuades not only if logical, but if it appears to be logical. If the persuasive power of logical linguistic argument were recognized as 58  De Man’s observations about the logic-grammar conjunction in the early Enlightenment are congruent with the uptake failure of Bacon’s sceptical views of language and Fish’s comments on the hegemony of the positivistic style. “Seventeenth-century epistemology, for instance, at the moment when the relationship between philosophy and mathematics is particularly close, holds up the language of what it calls geometry (mos geometricus) . . . as the sole model of coherence and economy. Reasoning more geometico is said to be ‘almost the only mode of reasoning that is infallible, because it is the only one to adhere to the true method, whereas all other ones are by natural necessity in a degree of confusion of which only geometrical minds can be aware.’ This is a clear instance of the interconnection between a science of the phenomenal worlds and a science of language conceived as definitional logic, the pre-condition for a correct axiomatic-deductive, synthetic reasoning. . . This articulation of the sciences of language with the mathematical sciences represents a particularly compelling version of a continuity between a theory of language, as logic, and the knowledge of the phenomenal world to which mathematics gives access” ( 13-14).  50  its allure, we might understand reason quite differently. In this logic-affect vein, Bacon highlights the charms of logical deduction: “Syllogism, as it is a thing most agreeable to the mind of man, so it hath been vehemently and excellently laboured” (AOL XIV). The pleasure of logic again recalls Aristotle’s appeal to feelings of confidence and fittingness. William James’ corroboration of the pleasures of logic helps to further disarticulate some of the linkages between language, logic and rational thought. In a gesture which seems to anticipate Tomkins’ views on the role of affect in cognition, James’ “The Sentiment of Rationality” argues that rationality is a sentiment – an affective response, “a strong feeling of ease, peace, rest.” A recently published study in experimental psychology over a century later echos James’ language when it identifies a “fluency-triggered positive affect” experienced “as a cognitive feeling of ease” (Topolinksi and Strack 40). The affective response identified in these experiments is provoked by intuitions of coherence very like the sentiment of rationality that James describes. James says that rationality is the experience of things fitting together, a “sufficiency of the present moment” and “absence of all need to explain” or to account for something (James 58). He says that “the transition from a state of puzzle and perplexity to rational comprehension is full of lively relief and pleasure” and suggests that there are good reasons for upholding the view that rationality is “the absence of any feeling of irrationality” (57). Coherence, for James, is experienced as a feeling; the sentiment of rationality is a feeling of cognitive “fluency” (58). Coherence and fluency – both associated with intuition in contemporary experimental psychology studies – show that affect plays a variety of roles in cognitive processes.59 For example, Topolinski and Strack list the range of studies in experimental psychology which have established that long-term, positive-mood affect facilitates “semantic spread,” increasing both  59 Bowers, et al., define intuition as “a preliminary perception of coherence (pattern, meaning, structure) that is at first not consciously represented, but which nevertheless guides thought and inquiry toward a hunch or hypothesis about the nature of the coherence in question” (74). Following this model Baumann and Kuhl “define intuition in terms of the perception of coherence” – the “spontaneous or preliminary perception of coherence that is not consciously represented” (1213). This definition together with the studies of the primary role of affect in intuition corroborate James’ original insight into the sentiment of rationality as well as directing attention to the question of part whole relations imbricated in cognition (for Van Dijk and Tomkins) and classification (for Plato and Wittgenstein) as we shall see. Both cognitive and classificatory uses of partwhole relations are linguistically (or semiotically) indexed.  51  accuracy and fluency in cognition. Their own efforts to refine a model of intuition determines that short-term, positive “core affect” contributes to conclusions of coherence (whether or not coherence exists).60 These findings, which both substantiate Tomkins ideas about the involvement of affect in cognition, and summons James’ wordings, further challenge the characterization of human thought – often called “reason” – as logical (or rational) . James’ treatment of rationality – as a kind of confidence (that things fit together) and a kind of feeling – is corroborated by these experimental-psychology studies which find that high fluency intuitions of coherence61 trigger positive affect. Topolinksi and Strack determine that the short-term “core affect” is the outcome of the intuitive chain triggered by fluency and, most significantly, that this core-affect response can be manipulated to produce counter-to-fact intuitions (that is, incoherent things are perceived as coherent) (58). In other words, the properly timed production of affect can make the logical appear unlogical and vice-versa – notwithstanding that the term “logical” (applied to linguistic usage), like its counterpart “rational” has lost some of its lustre in the course of this discussion. Rhetorical theory has known for a long time that when it comes to linguistic arguments logical appeal is not the syllogism, not the long train of reasoning, but the appearance or feeling of logic (see, Rhetoric 1.2.14). Perhaps an orator’s success is knowing how to trigger intuitions of coherence; and perhaps it is just such knowledge that provoke Socrates’ worries (in the Gorgias) that a medical patient might be more easily persuaded by a rhetor than a physician. Aristotle’s shrewd advice that a speaker target the audience to whom he speaks has some relation to James’ view that rational sentiments are stimulated differently in different individuals. Crudely stated, James suggests that some individuals are attracted to certainty like simple religious or logical/scientific accounts – whereas others crave excitement and are drawn to puzzles, difficulty and irrationality. While James is concerned with individual subjectivities, 60 Baumann and Kuhl conclude that “access to coherence producing cognitive processing . . . depends on affect regulation” (1221). According to Topolinksi and Strack this and other Kuhl studies advance an “affect modulation hypothesis” in which a general and lasting mood affect promotes or inhibits semantic spread and thus conceptual fluency (49). See Topolinksi and Strack’s extensive list of studies regarding mood affect and certainty judgements, risk taking, etc. (48). 61  In most of the studies referred to here semantic coherence was tested using word triads. This makes these studies particularly apt for consideration of the relationship of mathematical logic to language use.  52  the malleability of what is considered rational or coherent is also cultural and thus applicable to any community to whom Aristotle’s orator might be speaking (e.g., a religious community, a legal community, an ethnic community, an adolescent community). But James’ individual focus offers particular insight into the resistance to theory. Since the sentiment of rationality is produced by a complex of dispositions and these dispositions are different in different individuals, what in some, produces anxiety and creates an obstacle, for others stimulates curiosity and acts as enticement. If affective elements cannot be severed from coherence, rational thought must be both affectively modulated and guided by subjective disposition. Affect not only plays a part in a sign’s history, but language use itself is affectively motivated according to both Tomkins and the findings of scholarship on reported speech. Notions of explicit reference, logically presented description, and coded communication are now completely muddled, not only by the whimsies of the sign, but by the role of affect in human thought and motivation. Language is not a sterling instrument awaiting its proper use: a clean tool, incisive and clear-cutting. As a product of affectively conditioned experience, it is itself a musty medium. 2.5  epistemological quagmire: analogy & knowledge ~  as Plato saith, Whoseover seeketh, knoweth that which he seeketh for in a general notion; else how shall he know it when he hath found it? (Bacon AOL 48) While the wobbly condition of the sign renders linguistic mediation problematic in its own right, the role of affect in human rationality and the intercourse between affect and sign compound its already debilitated condition. The question of what constitutes rationality in the first place – what kind of mental operations are in play when we attempt to understand the world – might be asked. According to James, one powerful player is analogy, the cognitive mechanism through which identification takes place. When identification (using a linguistic sign) takes place by means of analogy to another thing, then the history of the sign (of the other thing) comes along for the ride. That “analogy lies at the core of human cognition” is a  53  commonplace of psychology and cognitive science.62 Green et al., say that “analogical thinking is a fundamental cognitive tool for learning, understanding and generating novel ideas” (1005). James’ illustration of how analogy stirs the sentiment of rationality highlights both the recognition element of analogical reasoning and the affective processes entailed in it. Citing Bain’s comments James notes that: “‘A difficulty is solved, a mystery unriddled, when it can be shown to resemble something else; to be an example of a fact already known.’” He adds that “the craving for rationality is appeased by the identification of one thing with another” (62-63). Here, the identification of a thing by analogy to another thing produces the sentiment of rationality: the feeling of satisfaction that comes from identification provokes coherence intuitions. If this is what human reason is (a feeling of things fitting together), and logic in linguistic argument is what we feel follows (what seems coherent), we might want to be circumspect about what kind of identification results from an affectively motivated analogy between thing A and thing B – especially when thing A is identified (not by a unique sign “A” but) by a sign over which it has no propriety in the first place, a sign any other thing can use. It is not just a question of affective involvement in rational cognitive functions, but a question of linguistically mediated identification (and classification) which makes all classified things caught up in the endless chain of signification. And it is not just the role of affect and semiotic promiscuity that are problematic, but the complex and error-prone nature of the cognitive use of analogy.63 French’s observations about the cognitive uses of analogy underscore the selectivity involved in linguistically-mediated, analogical identifications. He highlights that there are a variety of uses for analogical reasoning (that is, different types of cognitive functions like the generalizing-classifying function, or the unique-identification function). French’s list of analogy-uses includes, for example, 1) emphasis of particular selected  62 Holyoak cited in Leech, Mareschal, & Cooper. See “Analogy as relational priming: A developmental and computational perspective on the origins of a complex cognitive skill,” and responses in Behaviour and Brain Sciences. 63  Recall Foucault’s critique of classifying diseases according to symptoms in Birth of the Clinic as though symptoms indicate absolute genus-species categories. Under the heading “The Principle of Analogy” he discusses both the process of classification in order to identify diseases to start with and then the function of analogy within that process. He says, “it is the analogy of these relations that makes it possible to identify a disease in a series of diseases” (100).  54  features or parts of a thing, 2) generalizing a thing, 3) specifying a thing.64 But observe that reductive (generalizing) interests, particularising interests and interests in only one part, are not the same and will lead to different analogies for the same thing (that is, A, B, and C will all be used to identify D). Furthermore all of the analogies will be characterized by varying (and unmeasurable) degrees of inaccuracy depending on the context. Such comparisons are vague at best. French’s example of an American mistaking a European bidet for a toilet is a typical (if embarrassing) identification error. One wonders what kind of identification this analogical process could possibly produce and thus how “accurate” referentially descriptive code could possibly be – were it even remotely possible for the human brain to remember the number of terms that would be required to label everything determinately. Insufficient knowledge of the old (the toilet) or the new (the bidet) signals the limits of human perception and knowledge, leading to category mistakes and false identifications. “Analogy-making,” according to French, “is typically conceived of as involving a ‘mapping’ between two domains” (200). Analogy is a metaphorical operation, (meta-fevrw) a carrying over or transfer from one sphere to another (as Nietzsche says), neither precise nor absolute. What is classified as one thing might later be “discovered” to be something else as Kuhnian caveats warn and Plato’s Timaeus illustrates. Linguistically mediated identification and classification’s dependence on analogy renders these cognitive operations vague and fuzzy at best. Insight failures, misunderstandings, false assumptions make reasoning by analogy error prone and tending to category mistakes. But the expression “category mistake” itself is problematic; it wrongly implies the potential for “correct” identification. The term “correct” isn’t really applicable in reasoning by analogy. In his analysis of insurance claims, for example, Whorf observes the hazards posed by gasoline drums that were erroneously deemed “safe” because labelled “empty” – when instead they should have been deemed potentially more volatile than full ones (135). But the term “empty” elicits the category mistake “safe” with novice gasoline  64 French states that analogy making “encompasses our ability to explain new concepts in terms of already-familiar ones, to emphasize particular aspects of situations, to generalize, to characterize, to explain or describe new phenomena, to serve as a basis for how to act in unfamiliar surroundings, and to understand many types of humour” (200).  55  handlers. It seems to mean that the dangerous substance is gone and the danger absent, whereas it actually means the opposite, that the drum might now be “full” of an air/gasoline mix that is much more volatile. French refers to several abstract processes necessary for analogical reasoning including recognition of a source and target, mapping between the two, and transfer from source to target (200). These mapping and transfer processes are the essence of metaphor, of the tropological, so it is not just the linguistic sign but human reasoning itself that is tropological. “Recognition” (of a source and target), however, requires knowledge of a thing and careful inspection – not the overhasty and vague generalizing that Bacon worries about. In French’s example, the drain on the perceived “toilet” needed to be carefully apprised. Note also that the sign “toilet” was known by French, but the sign “bidet” was likely not known. James has suggested that rationality is triggered by “the identification of one thing with another” (like a bidet) as “an example of a fact already known” (like a toilet) (63). This recognition is conditioned by the paradoxical knowledge-recognition paradigm – namely, that one cannot recognize what one does not know – a salient insight, as Bacon reminds us, of Plato’s Meno. Wittgenstein too says: “One has already to know (or to be able to do) something in order to be capable of asking a thing’s name” (§30). One must be able to identify or delimit the boundaries of a thing in order to inquire of its name. Analogical thinking depends on recognition, which depends on (linguistically mediated) “knowledge” – a slippery commodity at best. And we can never know whether it is correct or incorrect, only effective or not effective; “accuracy” can only be estimated from empirical results. The cognitive use of analogy, and identification by recognition imbricate fuzzy, knowledge-dependent cognitive functions with the mediation of the linguistic sign. These fuzzy metaphorical operations call into question the degree of precision that is possible with a socially-conditioned and promiscuous linguistic sign. A finite lexicon together with limitations on human memory and perspective further alienate prospects of linguistic precision. But whence comes the prestige of precision? According to Wittgenstein, it is logic, not language use or  56  communication, that demands precision. He thematizes the often vague uses of language in terms of “exactness” and concept-delimiting “boundaries.” In a discussion of the whimsical linguistic operations of unregulatable language games, he queries “does it take that [a clearly delimited boundary] to make the concept usable? Not at all!” (§69). His Cratylian interlocutor later asks “But is a blurred concept a concept at all?” implying that precision failure means that the thing referred to does not exist. Wittgenstein dismisses this suggestion out of hand with the rhetorical questions: “Is an indistinct photograph a picture of a person at all? Is it even an advantage to replace an indistinct picture by a sharp one? Isn’t the indistinct one often exactly what we need?” (§71). “[Inexact],” he asserts, “does not mean ‘unusable’” (§87). And it seems possible that a generalized understanding of how language works might be more broadly usable than a precise one: a fuzzy picture of how language works, might be more serviceable for keeping linguistic mediation in mind, for keeping linguistic consciousness at the intellectual scene. Querying the prestige status bestowed by precision (and required only for logic) Wittgenstein goes on to discuss “the use of words with games and calculi” (§81): [I]f you say that our languages only approximate to such calculi you are standing on the very brink of a misunderstanding. For then it may look as if what we were talking about were an ideal language. As if our logic were, so to speak, a logic for a vacuum. – Whereas logic does not treat of language – or of thought – in the sense in which a natural science treats of a natural phenomenon . . . (§81) Wittgenstein argues that the precision of logic is not applicable to the kind of (non-ideal) language that we actually use (“ordinary language”) and which, he is at pains to point out, is perfectly orderly without emendation to propositional form. Halliday likewise observes that “the ‘grammatical logic’ of a natural language is different from mathematical logic” (285). Wittgenstein’s move – displacing the paramount role that had formerly been reserved for mathematical logic in linguistic communication – is significant for analytic philosophy. The codification-resistant system of language games Wittgenstein uncovers cannot be regulated in  57  the determinate way that some philosophers (like Wittgenstein himself) had once hoped.65 2.6  fuzzy logic and the biochemical image of grammar ~  Whatever its failings the state of scientific technology is a persuasive indication that language works. Human reasoning processes use some kind of relatively “logical” process despite linguistic-analogically-produced fuzziness and even if we cannot delimit the precise extent to which intuitive-logical principles or affective responses govern specific processes and their outcomes. Sperber and Wilson’s conception of the role that deduction plays in nondemonstrative inference (that is, in inferences where conclusions are not necessitated by premises) seems also applicable to what we more broadly think of as rational thought. Recall that “logical” is one of Halliday’s metafunctions associated with ideational processes and parallel to the experiential dimension which is construed by grammar. In the ostensive inferential model, deductive functions are limited by a more or less secondary (but crucial) role played in producing contextual implications from utterances (the effect produced from the combination of what is already known and the assumptions that have been made manifest). This “deductive device” explicates the contents of assumptions according to a set of formal elimination rules together with other non-deductive heuristics to produce contextual effects. Contextual effects are then gauged according to relevance: a measure of their size and their processing effort that is a kind of informed-intuitive hunch. Sperber and Wilson speculate that traditional, logical, elimination rules (modus ponens, modus tollens, conjunction, disjunction, etc.) spontaneously performed in non-demonstrative inference “might be less a logical process than a form of suitably constrained guesswork” (69).66 Although the entire enterprise would be qualified by the un-delimited role of affect (that is, of motivation) and the fuzziness of linguistic-analogical-reasoning, results can be quite “logical” in a fuzzy sort of way, just not as precisely so as the logical positivists and analytic philosophers may have wished. While Sperber and Wilson’s speculations about the role of a deductive device in 65  We see the way that the Tractatus provides context to the Investigations when Wittgenstein says: “it will then also become clear what may lead us (and did lead me) to think that if anyone utters a sentence and means or understands it he is operating a calculus according to definite rules” (§81). 66  They say: “human spontaneous non-demonstrative inference is not, overall, a logical process” (69).  58  producing contextual implications from utterances are focussed specifically on the mechanisms of communication, they assert that generally “humans are rather good at non-demonstrative reasoning; otherwise the species would be extinct” (68). This takes us back to the question about what motivates reason. Since the premises, for Sperber and Wilson, do not necessitate the conclusions, non-demonstrative reasoning might look a lot like narrative. Whatever logic and rationality are exactly, it seems clear that commonsense assumptions about them need to be questioned. We need to keep in mind reasoning’s cognitive dependance on analogy (or metaphor) and rationality’s motivation and (affect-laden) linguistic ties. Rather than telling us something about how language works, Wittgenstein claims that logic is an indicator of human mental habits. He says “the laws of logic are indeed the expression of ‘thinking habits’ but also of the habit of thinking. That is to say they can be said to shew: how human beings think, and also what human beings call ‘thinking.’”67 This suggestion resonates interestingly with Halliday’s conception of grammar as we shall shortly see. The questions arising here include what criteria determine the logic of these thinking habits, and what produces linguistically formulated patterns of logical entailment. Fahnestock takes up the question of rational “argument” in the writings of early modern science. She analyzes these texts to determine how their writers thought about science and concludes that the models of reasoning evident in the writings of Enlightenment science were produced by the rhetorical-dialectical training of the day. She compares the exhaustive forms of argument catalogued in sixteenth-century-humanist dialectic manuals (which were a centrepiece of early modern education applicable in all disciplines) to the science writing of the Enlightenment.68 These manuals provided named rhetorical figures and patterns of argument, memorized and practised by students; “the pattern itself,” Fahnestock claims, was considered “the source of its ‘reasonableness’” (Stylistics 226). Inscribed in the reasoning processes of  67 See Remarks on the Foundations of Mathematics (§ 131). He adds “the propositions of logic are ‘laws of thought,’ ‘because they bring out the essence of human thinking’—to put it more correctly: because they bring out, or shew, the essence, the technique of thinking. They shew what thinking is and also shew kinds of thinking” (§133). 68  In Figures she analyzes the arguments of Kekule, Mendel, Lavoisier and Harvey who write on matters falling into modern disciplines of chemistry, genetics, palaeontology, medicine.  59  Enlightenment-scientific texts she discovers the contours of dialectical arguments from the training manuals. Arguing that the correspondence between the reasoning and the composing practices of these writers stemmed from their dialectical training, she concludes that what passes for rational thought was shaped by Renaissance dialectical pedagogy. Jardine’s analysis of Bacon’s art of discourse makes clear that what theorists like de Man call rhetoric is what Bacon treats as dialectic. Fahnestock clarifies that during this period dialectic and rhetoric were in such close proximity as to be barely distinguishable.69 Thus both terms are associated with the practice of argumentation that Fahnestock analyzes as well as the verbal patterns she calls “figures of argument.” It seems very likely that Wittgenstein’s “thinking habits” that constitute the “laws of logic” might line up nicely with the verbal patterns inscribed in dialectical training manuals. Fahnestock refers to the (now lost) “intimate connection” between figures and topics in the training manuals’ hybrid disciplinary approach (Stylistics 228, 224). The argumentative strategies Fahnestock researches (which are adumbrated under both style and invention categories in rhetoric-dialectic manuals) include classical rhetoric’s topoi: generally understood as lines of argument (and hence invention). Perhaps this style-argument ambiguity explains why contemporary theorists have found Aristotle’s topoi to be “confused” and why there is no agreement about exactly what Aristotle means by “topos.”70 According to Leff the ambiguity about what topoi are remains constant throughout antiquity.71 Leff observes that the “term ‘topic’ incorporates a bewildering diversity of meanings” and that “among modern authors we find conceptions of the topics ranging from recurrent themes in literature, to heuristic devices that encourage the innovation of ideas, to regions of experience from which one draws the substance of an argument” (Topics 23). Whether as recurrent themes, invention heuristics, or  69  Fahnestock cites Melanchthon’s early sixteenth century comment: “So great is the resemblance between Dialectics and Rhetoric that scarcely any distinction can be observed between them” (Figures 122). 70 See for instance McAdon Signs, Rhetoric, Leff Topics, Topoi For related discussion of invention see Miller Special Topics, Novelty. 71 Leff says that “the number of classical conceptions of topics rivals the Babel of modern scholarship, since classical rhetoricians are not very scrupulous in defining this key term. Aristotle, for example, makes the topics one of the central elements in Rhetoric without ever explicitly defining what they are” (Topics, 23).  60  patterns and materials of pre-existing argument, curiously enough there is widespread agreement on the topoi’s strictly pedagogical purposes – a view which Quintillian makes unquestionably clear (see Leff Topics 34). Topoi were designed for practice exercises not as ends in themselves; they taught students how to think, that is, they taught verbal patterns of reasonableness. Leff says that “just as gymnastic training prepares the body for physical performance, so can exercises in topical reasoning prepare the mind for verbal performance” (Topoi 208). The suggestion that what is “logical” or rational (perhaps stemming as Wittgenstein suggests from thinking habits) might be tied to the verbal patterns of linguistic usage lines up with Halliday’s conception of grammar. There is no question that for Halliday thinking habits are embedded in the grammar of language, in what he calls the “wordings,” representations that organize meanings in lexicogrammar. Halliday believes that our “extraordinarily rich and varied” human experience is “mediated through the senses on various levels” and that “what the grammar does is to transform this experience into meaning” (Matter 63). Another way of saying that “grammar ‘transforms experience into meaning’” is to say that grammar is a theory of human experience. The lexicogrammatical representations or wordings (I should make clear that ‘grammar’ includes vocabulary, the more specific elements of the wording) – that is, the bits of discourse that we recognize as speech, or as writing, construe fragments of experience, real or imaginary, in ways which relate each instance to this overall theory. (Matter 63-64) The wordings or bits of discourse theorize how the world works. The construal of experience constitutes a “theorizing [of experience] in the form that we call understanding” and sets up “logical-semantic relationships between one clausal unit and another” (OL16-17). These (grammatically) “logical” relationships learned through language are indeed an expression of human thinking habits, as Wittgenstein says. Halliday’s account suggests that grammar is syntax and vocabulary, learned patterns of speech, wordings, bits of discourse: that is, very like Fahnestock’s verbal patterns derived from rhetorical figures – the “figural logic” of verbal topoi, received, recognizable lines of argument, more like narrative than (deductive-  61  mathematical) rules. Recall de Man’s comments affiliating “logic” and rule-governed grammar. Commonsense notions of grammar circulate in code-models of communication and conceptions of rational thought: where logic is imagined as symbolic/mathematical formula, and grammar as logic’s mirror image in linguistic signs. In this commonsense image, grammar is a set of logical rules to be followed. Halliday’s radically different understanding of what grammar is specifically excludes such conceptions: grammar is “not, of course, the image of ‘grammar’ that many people carry around with them” which is “the grammar of primary school, where they were taught that a language is a set of rules to be obeyed” (Matter 62). For Halliday, who claims that the organizing concepts of his thought emerged as a by-product of struggling with particular problems in language, “a language is a system of meaning – a semiotic system” (OL 1-2). “A language is almost certainly the most complicated semiotic system we have,” he says; “it is also a very fuzzy one, both in the sense that its own limits are unclear and in the sense that its internal organization is full of indeterminacy” (OL 2). He writes that “the grammar of any language can be represented as a very large network of systems” but that “the network is openended” (OL 180). For Halliday systemic theory is “like language itself”: a “system whose stability lies in its variation”; it is “a ‘metastable’ system,” which “persists because it is constantly in flux” (OL 192). “A lexicogrammar is not a closed, determinate system” and clearly has little in common with rigid conceptions of logic as an absolute grid of entailments according to determinate rules (OL 194). Halliday’s “five critical features of the grammar of a natural language” present an entirely other image of grammar than the notably-absent adjective “rule-governed” conveys; highlighting the flexibility and fluidity of language, systemic grammar is said to be “comprehensive, extravagant, indeterminate, non-autonomous and variable” (OL 250). The construal of experience in language not only creates meaning, but for Halliday (as for Whorf) “grammar models reality” (OL188). Language is used to “construe experience, where construe means ‘construct semiotically’ – that is transform experience into meaning”; Halliday explains  62  that “experience [here] relates the process of construal to the experiencer rather than the ‘real world’” (OL 244). We will return to these shaping, constraining/enabling features of grammar on our interpretation of the world in Part 3, but for now, it is important to recognize the twosided, symbiotic role of lexicogrammar. For Halliday language both is meanings (semiotic) and makes meanings (semogenic). He says the usual way we talk about language is by saying that language ‘expresses’ meaning, as if the meanings were already there – already existing, in some formation or other, and waiting for language to transpose them to sound, or into some kind of visible symbols. But meaning is brought about by language; and the energy by which this is achieved, the source of its semogenic power is grammar. (Matter 63) Halliday here echoes Foucault’s notion that language produces what it seems to describe and his grammar, figured here as energy, resembles the complex dynamics of biochemistry more than any set of “logical” rules we could use to analyze truth conditions of propositions. Fahnestock seems to be thinking along the same lines as Wittgenstein and Halliday when she suggests that “reasoning and [verbal]-pattern completion” are in close proximity (Figures 124). When he asks “if language is a code, then where is the pre-coded message?” Halliday is also thinking along the same lines as Voloshinov of the liveliness of semiotic sphere and of grammar (OL 243). We are reminded that thinking takes place in living language not abstract mathematical symbols. Halliday’s innovative view of grammar as wordings, also resonates with Voloshinov’s account of the fragmentary nature of inner speech and human linguistic thought. In Voloshinov’s account thinking, or inner speech, employs signs or words, not in sentences, but in fragmentary utterance that are “joined together” and “alternate” not according to received commonsense notions of the “laws of grammar or logic, but according to the laws of evaluative (emotive) correspondence”: The units of which inner speech is constituted are certain whole entities somewhat resembling a passage of monologic speech or whole utterances. But most of all, they resemble the alternating lines of a dialogue. There was good reason why thinkers in ancient times should have conceived of inner speech as inner dialogue. These whole entities of inner speech are not resolvable into grammatical elements (or are resolvable only with considerable qualifications)  63  and have in force between them, just as in the case of the alternating lines of dialogue, not grammatical connections but connections of a different kind. These units of inner speech, these total impressions of utterances, are joined with one another and alternate with one another not according to the laws of grammar or logic but according to the laws of evaluative (emotive) correspondence, dialogical deployment, etc., in close dependence on the historical conditions of the social situation and the whole pragmatic run of life. (38) Alternating lines of dialogue are not logical propositions, but bits and pieces – like Halliday’s “bits of discourse.” As we shall see in Part 3, the transcripts of actual conversational dialogue used in studies of reported speech show that communication is often effected ostensively in gestures – using single words of support, repeating what another has just said, play acting someone else’s utterance – rarely in the whole sentences of a rule-governed grammar. Again, as if he were in a conversation with both Fahnestock and Halliday, Bolinger asks “is grammar something where speakers ‘produce’ (i.e. originate) constructions, or where they ‘reach for’ them, from a preestablished inventory, when the occasion presents itself?” (381).72 Voloshinov is mindful of the historicity of the sign with its constant re-accentings and accretions when he draws attention to attendant stylistic shifts that effect what he calls gramaticalization and degrammaticalization. He says that “stylistic individualization of language in concrete utterance is historical and creatively productive. It is here precisely that language is generated, later to solidify into grammatical forms: everything that becomes a fact of grammar had once been a fact of style” (51). Voloshinov seems here to imagine bits of discourse or wordings that begin life as a stylistic innovation and then become established norms. He says later that “the borderline [between style and grammar] is fluid because of the very mode of existence of language, in which, simultaneously, some forms are undergoing grammaticalization while others are undergoing degrammaticization” (126). Recall Fahnestock’s account of the overlapping classifications of argument and style: fluid boundaries bridging style, syntax, patterns of speech and argument she calls figures of argument. With his mind on rules established after a stylistic innovation has become an established norm, 72  “Remembered prior texts” is a stock phrase in Becker: “the actual a-priori of any language event – the real deep structure – is an accumulation of remembered prior texts” (86).  64  Voloshinov denounces the philological focus of early-twentieth-century linguistic study, claiming that “systematizing grammatical thought could have developed to its full scope and power only on the material of an alien, dead language” (78).73 Such systematizing (rulegoverned) grammatical thought he says, “interpret[s] living language as if it were already perfected and ready-made and thus must look upon any sort of innovation in language with hostility” (78). Voloshinov suggests that beneath views of “language as a system of normatively identical forms lies a practical and theoretical focus of attention on the study of defunct, alien languages preserved in written monuments” (78,71). Halliday too observes this “historical association of linguistics with writing”; he observes that “linguistics begins when a language is written down, and so made accessible to conscious attention”; and he comments on the distorting influence of what others have called the written language bias, to which we return in Part 3 (OL 193). Voloshinov’s focus on the sign’s historicity and his celebration of the sign’s fluidity recognize that linguistic semiotic materials, their grammatical forms and stylistics, are fluid, constantly reacting, responding and changing. Halliday offers a new way of thinking about grammar, that welcomes the caprices of Voloshinov’s living historical sign, and an image that allows us to glimpse the complex biochemistry of grammar – too multifaceted, flexible and vast to be a set of memorisable rules. This image of grammar invites reconsideration of rational thought as verbal patterns and syntax of language that we learn from our parents and society – or perhaps from Renaissance-dialectic manuals. Halliday’s notion of grammar has fuzzy logic: it is too complex to keep track of as a set of “rules” to be followed and the grammar of individual words is often unique.74 Grammar is mutable like language itself. Such fuzzy linguistic logic may have natural ties to the standard elimination rules that Sperber and Wilson describe. Though it may have less prestige than formerly attributed to the mythical-precise linguistic logic, their non-demonstrative deductive guesswork seems to work quite well.  73 74  A “system,” says Halliday, “is not the same thing as ‘systematic’” (OL 180).  Wittgenstein says “The grammar of the word ‘knows’ is evidently closely related to that of ‘can,’ ‘is able to.’ But also closely related to that of ‘understands’”(§150)  65  Nevertheless, Cratylian code-metaphor convictions of a rule-governed grammar still prevail in accounts of how language works. Sperber and Wilson complain that “despite the widespread scepticism about the role of deductive reasoning in comprehension, many existing pragmatic theories, especially those built on Gricean lines, seem to be based on informal systems of just this type” (93). 2.7  epistemological quagmire: part-whole relations ~  The human language-thought ligature together with rational thought’s affective components and analogical cognitive operations are further complicated by the subjective vantage of human language users. When it comes to epistemological identification and classification efforts, our perspective limitations compound the problems already amassed by the delinquent habits of the sign. To begin with, Cratylian conceptions of rule-governed grammar do not explain how human memory could manage the number of terms necessary for the improbable-impossible undertaking of uniquely identifying every aspect of reality. However, Van Dijk’s work on the cognitive functions of classification and specification helps us understand how human memory makes do with a finite lexicon and linguistic signs. James has anticipated the trajectory of Van Dijk’s cognitive functions when he distinguishes a reductive “passion for parsimony, [and] for economy of means in thought” (which is a focus on generalization or abstraction) from a “passion for distinguishing” which is an “impulse to be acquainted with the parts rather than to comprehend the whole” (58-59). Identification of a part and its classification according to category are specification and generalizing techniques that Van Dijk identifies in his work in linguistics, cognitive psychology, and discourse theory. Considering foregoing meditations on affective motivation and sentiment in rational thought, and linguistic complexity more generally, it should not be surprising that Van Dijk’s explorations of the workings of language found it necessary to resort to cognitive-functional explanations. His analyses determine that specification and generalization techniques are fundamental, paired cognitive functions used for memory and cognition. He describes these discursive  66  cognitive operations (of specification and generalization) as global structures that organize local information for storage, retrieval and comprehension. These generalizing (or abstracting) and specification (or particularizing) cognitive operations depend on “an intuitive primitive” function of part-whole relations which Van Dijk asserts “cannot be analyzed into more basic cognitive notions” (3). Generalizing and specifying are then basic cognitive operations conducted in linguistic signs. An epistemological problem is posed by identification and classification however: linguistically framed identifications by analogy depend on being able to see, in the figurative sense (that is, knowing), what part fits into what whole – and there are many wholes in which a given part might fit. This brings us back to the question of context. Of course such fitting is a human passion; fitting parts together into a whole induces the sentiment of rationality. But how can we know that any given thing fits into any particular context, and how that context fits into the larger context? When the different contexts into which a thing seems to fit are mutually exclusive (a problem guaranteed by the linguistic sign’s promiscuity), when the same word is used to describe radically different things (that cannot be classified together) or to describe things that can only be classified together in limited ways (but are otherwise too different), the limited perspective of human vantage interferes with epistemological identifications. Arguably the total knowledge-vision (omniscience) required to know where to draw these lines is superhuman and requires divine speech. The Greek term qewrei:n, the etymological root for ‘to theorize,’ means ‘to see,’ ‘look at,’ ‘view,’ ‘behold’ (often in a public or sacred context).75 Recall that when he is describing the requirements of the art of rhetoric Socrates says “I myself am a lover of these divisions and collections, that I may gain the power to speak and to think (emphasis added), and whenever I deem another man able to discern an objective unity and plurality, I follow in his footstep where he leadeth as a god . . . for those who have this ability [and if Socrates doesn’t who does?] . . . I  75 Gonzalez raises this vision-beholding interpretation of qewrei:n in connection with Plato: “what one gains through entering the world of dialogue is not a ‘theory’ in the sense of a ‘doctrine,’ but a ‘theory’ in the sense of a ‘vision’” (Third 18). See also Statkiewicz (2).  67  call dialecticians” (Hackforth Phaedrus 266b).76 Dialectician, here, has nothing to do with Hegelian dialectic (thesis or argument); the root terms for Platonic dialectic and dialectical are built on diav (through) and levgomai (‘speak,’ ‘say’); the parts of dialectic add up to something like ‘through speech.’77 Plato’s dialectician is s/he who is skilled in speaking (and thinking) and s/he who can through speech discern a collective unity and plurality. It is this impossible knowledge and ability that are needed to make accurate identifications and classifications – impossible knowledge and skill that warrant the Socratic stance of ignorance. This inaccessible qualification is required for speaking and thinking in Plato’s art of rhetoric. Only omniscient divine speech (with infinite memory) could be code. Wittgenstein points out that one must be able to determine the “simple” from the “composite”– an always shifting part-whole determination that is thematized in Philosophical Investigations. Lacking divine speech and the vision of a dialectician, humans range from parts to wholes (never knowing when they have been misled by the promiscuities of the sign). The problem of where one whole becomes a part and what delimits the parts and the wholes – the interminable question of where to draw the line that preoccupies Burke – always blinkers identification. Wittgenstein asks, “if it had been laid down that the visual image of a tree was to be called ‘composite’ if one saw not just a single trunk, but also branches, then the question: ‘Is the visual image of the tree simple or composite?’ and the question ‘What are its simple component parts’ would have a clear sense” (§47). Implicitly, it depends on what constitutes a whole (in this instance, an image of trunk and branches) and what constitutes a part (in this instance he refers only to the trunk and branches). If the scope is predetermined then the question about what the composite whole is and what parts it has “would have clear sense.” Distinguishing always 76 Aristotle’s Posterior Analytics attempts to deny this problematic: “In defining by division there is no need to know all the facts. Some, however, maintain that it is impossible to know the differentiae between each thing and the rest without knowing each thing severally, and impossible to know each thing severally without knowing the differentiae” (Post. An. 2.13, 97a).  According to Schiappa, dialektikhv (the art of speech) is one of several terms Plato coined by adding the term tevknh to an existing root including rJhtorikhv or the art of rhetoric (skill, art, craft) (15). These two terms can have a very similar meaning and so called “good” rhetoric and dialectic are aligned in Phaedrus. The ability of the dialectician is what is needed for the “good” rhetorician; the “bad” rhetorician does not have this ability/knowledge. Socrates says: “Then the art of contention in speech in not confined to courts and political gatherings, but apparently, if it is an art at all, it would be one and the same in all kinds of speaking, the art by which a man will be able to produce a resemblance between all things between which it can be produced, and to bring to the light the resemblances produced and disguised by anyone else” (Loeb 261e). 77  68  recursive part-whole relations at any particular moment (in any particular context) is the unending and inevitable challenge in any identification and classification project, the Parmenidean question of where to draw the line. Bacon addresses this matter in terms of style – a fact of some consequence to this thesis’ stylistic focus.78 He complains about those who attempt to transmit sciences “as if they were complete in all parts and finished,” suggesting that their method and divisions pretend to “present the form and plan of a perfect science.” But he suggests that “their divisions are ill filled out and are but as empty cases.” The part-whole, identification-and-classification difficulties presented by these empty cases are a reminder of the post-modern concern with the violence of the line with its long history dating back to Parmenides’ views about the unity of reality.79 Where ever the line is drawn it necessarily covers up what is under it, often arbitrarily slicing up what is otherwise unified, and imposing a mark which becomes a lens through which it is seen. De Man suggests that insight is always attended by a degree of blindness; Parmenides’ monism prompts epistemological worries about the arbitrary lines drawn by linguistic signs; and Foucault complains in Birth of the Clinic about how the medical gaze carves up the scene. Broaching the same totalization problem that Bacon does about filling in the gaps Foucault observes that “total description . . . is much more the dream of a thought than a basic conceptual structure” (115). Halliday too acknowledges the problem of a language and a  78  “Further, this admiration of men for knowledges and arts — an admiration in itself weak enough, and well-nigh childish — has been increased by the craft and artifices of those who have handled and transmitted sciences. For they set them forth with such ambition and parade, and bring them into the view of the world so fashioned and masked as if they were complete in all parts and finished. For if you look at the method of them and the divisions, they seem to embrace and comprise everything which can belong to the subject. And although these divisions are ill filled out and are but as empty cases, still to the common mind they present the form and plan of a perfect science. But the first and most ancient seekers after truth were wont, with better faith and better fortune, too, to throw the knowledge which they gathered from the contemplation of things, and which they meant to store up for use, into aphorisms; that is, into short and scattered sentences, not linked together by an artificial method; and did not pretend or profess to embrace the entire art. But as the matter now is, it is nothing strange if men do not seek to advance in things delivered to them as long since perfect and complete” (Novum Organum LXXXVI). 79  Parmenides’ goddess characterizes her own language as duplicitous when she advises that we will learn by “listening to the deceitful ordering of [her] words” (8.52). The words are deceitful precisely because they are words, signs – distinguishing one thing from another. When language carves reality up into words (naming two forms where there is only one reality), it misrepresents what is one as many: Here I end my trustworthy account and thought concerning truth. Learn hence forth the beliefs of mortals, harkening to the deceitful ordering of my words. For they have made up their minds to name two forms, one of which it is not right to name – here is where they have gone astray – and have distinguished them as opposite in bodily form and assigned to them marks distinguishing them one from another. (emphasis added 8.50-56, Robinson 118)  69  grammar that necessarily makes a fluid world sit still – that it is “we who cut it up with our verbs and nouns into processes and things” (OL 244). Drawing together both the part-wholecomprehension problem (a vision-knowledge handicap) and the analogy problematic Halliday argues both that one could say there are “no natural classes in the perceptual world” and also that “there are far too many natural classes; phenomena resemble each other in indefinitely many ways, hence there are innumerable ways of categorizing, leading to variation not only between one language and another but also within one and the same language” (OL 244). Again it is a question of human motivation and what motivates where the line is drawn. One might pretend that they can describe everything in words – but since no one is omniscient, the ability to comprehend or see the whole and the parts is inaccessible. When it comes to looking at language this problematic is exponentially multiplied. Halliday writes: “the categories we set up to explain how language works are almost all inherently fuzzy”; the descriptive categories, themselves “fluid and unstable” constitute the “fuzzy sets” of “the categories of language itself”; these “categories themselves are inescapably the product of compromise, since the different perspectives locate the boundaries between categories at different places” (OL 28, 266). Thus systemic theory adopts what Halliday calls “trinocular vision” – at least a threefold perspective on linguistic categorization. He observes: Indeterminacy is bound to arise in language because the grammar is constantly juggling with conflicting categorizations, accommodating them so as to construe a multidimensional meaning space, highly elastic and receptive to new meanings. In doing so, the grammar adopts a kind of trinocular vision, giving it a threefold perspective on the categories and their configurations. (OL 254) I suggest that this category problem in language theory is paradigmatic of any classification undertaking because of the linguistic mediation of epistemological endeavours. Furthermore, just as logical and affective reasoning processes might be impossible to delimit and disentangle, some things might not be verbally analyzable in a useful or productive way. It is really a question of the limits of what we can see (i.e., the level of detail and the distance), what we can comprehend in a single glance (i.e., the penetration of human understanding and the scope of cognitive abilities), and the subjective vantage from which we  70  look. Perhaps a picture or image is worth more than a thousand words. As Wittgenstein says: Compare knowing and saying. how many feet high Mont Blanc is – how the word “game’ is used – how a clarinet sounds. If you are surprised that one can know something and not be able to say it, you are perhaps thinking of a case like the first. Certainly not one like the second. (§78) But what is of most interest here is the second case: try saying “how the word ‘game’ is used.” Arguably one could say many things about Wittgenstein’s language games. Indeed the list would be so long, complex, the elements interfused and at times contradictory, that arguably, apart from scholarly classification, their expression would be too copious and paradoxical for practical applications. Systemic grammar includes a family of related analyses – a complex network whose affinities and interrelations are doubtless only accessible to grammar specialists.80 This copiousness-complexity problem characterizes the generalized expressions of sceptical (eyes-wide-open) language theory. Quintillian’s comments on the list of topoi he devises as a culturally-specific training apparatus (with a best-before date) also speaks to the grammatical-codification-of-linguistic-operations problematic. According to Leff, Quintilian “disclaims any intention to devise a fully rigorous and exhaustive topical system. The effort would be futile and useless; it would cause the theorist to become prolix without being complete” (Topics 33). The interfused nature of the operations of language, the multiple aims of a single utterance, salient factors in linguistic complexity, recall the metaphor of the stream at the root of rhetoricity (rJevw – ‘flow’). What is of most interest to this thesis is how this aspect of linguistic complexity – language’s multifunctional, interfused nature, its ability to do several things at the same time – might determine genre and style choices of sceptical language theorists if only in  80  Halliday states: “the value of a theory lies in the use that can be made of it, and I have always considered a theory of language to be essentially consumer-oriented. In many instances the theorist is himself also and at the same time a consumer, designing a theory for application to his own task . . . Since there are so many tasks for which one needs a theory of language, any particular theory is likely to be, or very quickly become, a family of theories – still on speaking terms, one hopes, and with a personal rather than positional family role system. This is why there is no orthodox or ‘received’ version of systemic theory, such as may arise with self-contained systems that are impervious to influences from outside” (OL 192).  71  order to evade copiousness and paradox (the onerous risks of contradiction).81 Leaving completeness (totalization) to the divine order, Bacon, like Wittgenstein, opts for the stylistic practice of the “ancient seekers after truth” who wrote aphorisms; that is, “short and scattered sentences, not linked together by an artificial method,” and not pretending or professing “to embrace the entire art.” 2.8  the inherent resistance to theory: de Man and rhetorical theory ~  Referring to rule-governed grammar, de Man argues in “The Resistance to Theory” that “no grammatical decoding, however refined, could claim to reach the determining figural dimensions of a text” (16). In other words, it is figural-rhetorical features that finally qualify meaning and often, then, only in an indeterminate way; and this is not only because of the role of affective human motivation in language use and the sign’s tumultuous history, but rather language’s logic-resistance stems from its fundamentally tropological character. “Tropes, unlike grammar,” de Man asserts, “pertain primordially to language” (15). De Man illustrates the tropological character of language by producing a paradox in his own text. In order to show or perform the figurative and indeterminate nature of language he transforms the Cratylian resisting subject of his article into a sceptical resisting theorist: thus resistance is transformed from blindness to insight. Resistance had been thematized from the paper’s outset as a sign of psychological, ideological repression: a response to writing about writing by those whose Cratylian anxiety about uncertainty or whose vision-failure leads them to overlook sceptical language theory from Plato to Derrida. De Man claims that resistance to theory is a resistance to the rhetorical or tropological dimension of language: Rhetoric, by its actively negative relationship to grammar and to logic, certainly undoes the claims of the trivium (and by extension, of language) to be an epistemologically stable construct. The resistance to theory is, a dimension which is perhaps more explicitly in the foreground in literature (broadly conceived) than in other verbal manifestations or – to be somewhat less vague – which can be revealed in any verbal event when it is read textually. (17) 81  Halliday observes that “the extravagance of modelling that same domain of experience in more than one way leads to a richer and more life-supporting account. Of course, this will always leave room for what Claude Hagège (1997) referred to as “unheeded contradictions,” the leftover bits of language-building materials that continue to lie around; but the principle of contradictory construal is intrinsically a productive one” (OL 253).  72  “Literariness,” de Man says, is “the use of language that foregrounds the rhetorical over the grammatical and the logical function” (14). It is this awareness of the “literary” or rhetorical effects of language that creates language scepticism and sceptical (eyes-wide-open) theories of the linguistic sign. Thus de Man performs a figurative transformation (a turning) himself: resistance morphs from a psychological neurosis into a liberation front. Recognizing language’s inherent rhetoricity, resistance defends the embattled sign’s rights to promiscuity. Anxietydriven resistance is displaced by its opposite, an inherent resistence to theory. No longer a sign of repression resistance to theory becomes a sign of awareness: inherent in the idea of theory and its enterprise, inherent in the attitude of a sceptical-language theorist. What exactly is theory then that makes sceptical language theorists resist it on principle? We will return to the inherent inexactness of a linguistic semiotic that is fundamentally tropological with respect to Nietzsche in Part 3. But the question of inexactness, already raised by Wittgenstein, prompts consideration of whether a generalized language “theory” would be more useful than a precise one. Querying the precision required of ideal logical propositions Wittgenstein writes: “you still owe me a definition of exactness”(§69). Generalization (as a cognitive classification function) is, after all, a very useful and important linguistic cognitive function. One, albeit vague and abstract, definition for theory is “a conception or mental scheme of something to be done, or of the method of doing it; a systematic statement of rules or principles to be followed” (O.E.D.). By definition theory is general and conceptual. The Oxford English Dictionary definitions of “theory” include the collocates “scheme,” “method,” “systematic,” “rules or principles.” If we were performing coherence tests here we might think “grammar” (in sense of “rules”– even the arrangements of discourse that Hallidays’ systemic theory treats can also be described as variable and flexible rules, lexico-specific patterns). There is a part-whole relation between “theory” and “grammar” as rules. Grammar rules are general rules for particular instances of language use. A grammar is a kind of theory.82 And sceptical theorists – whose theories are codification resistant – are also theory-  82  Recall Halliday’s comment that “grammar is a theory of human experience” (Matter 63)  73  resistant because, according to Wittgenstein, theories can impede or blinker understanding. In his discussion of “exactness” (§88) – the (rule-governed) grammatical investigation of the “essence of language” (§92), and the “purest crystal” of logic “beneath the surface of language” (§97, 92) – Wittgenstein complains that “the crystalline purity of logic was, of course, not a result of investigation: it was a requirement” (§107). It was a requirement and an ideal that preceded investigation, acting as a lens. “Where does this idea [of the necessarily logical structure of propositions] come from?” he asks (§103). “It is like a pair of glasses on our nose through which we see whatever we look at. It never occurs to us to take them off” (§103). “One thinks that one is tracing the outline of the thing’s nature over and over again, and one is merely tracing round the frame though which we look at it” (§114). The idea of propositions having to be (mathematically) logical is an example of the way that a conception and a theory becomes a lens. This commonplace rhetorical insight – already commented on by Voloshinov,83 explicitly explored in Kenneth Burke’s article “Terministic Screens,” and treated in Halliday as the constraining and enabling operations of language – has its roots in Parmenides’ ancient complaint about the deceitful nature of words that divide a unified reality with the arbitrary marks of linguistic signs. Cratylian, code-metaphor theories of language were a requirement for the positive science that became the hallmark of the Enlightenment, thus becoming the organizing lens through which language was viewed. But Wittgenstein wants to empirically investigate how ordinary language actually works – not to project some idealized theory onto it. And de Man suggests that literary theorists want to know how textual effects are produced that “no grammatical decoding, however refined, could claim to reach” (16). If theorizing about how language works means finding a schema or overview, a systematic method for the operations of rules for how language works, then it is necessary to understand how parts are related to wholes. In the absence of divine memory and divine vision, the biochemistry of Halliday’s grammar presents an image of grammatology in which the rules are too copious and too complex to keep 83  Recall, Voloshinov says that “a sign does not simply exist as part of a reality – it reflects and refracts another reality. Therefore, it may distort that reality or be true to it, or may perceive it from a special point of view, and so forth” (10).  74  track of. When Wittgenstein empirically investigates how language works without the occluding lens of ideal propositions, he encounters a very messy and fuzzy linguistic complexity and a myriad of singular peculiarities with codification-resistant logics and topsy-turvy part-whole relations. What would it mean to systematize them? He claims not to be able to. And are not just a few of his examples worth at least a thousand words? Without being able to comprehend the collective unity and plurality (required of the dialectician for speaking and thinking), without the omniscience it takes to see all the partwhole relations in which linguistic semiotic materials are enmeshed, a systematic schema of how language works will always be as Halliday argues, indeterminate, open-ended, and in constant flux. Our intimacy with the linguistic sign doubtless contributes to our misunderstandings of it. The restricted selective human perspective cannot “behold” all the partwhole relations needed to trace the schema at any given moment. In the face of such immense complexity, the awareness prerequisites of one’s own vision limitations, semiotic dependencies, and cognitive deficiencies warrant the Socratic stance of ignorance. If theorizing language means to map the coordinates of how language works in a coherent, self-consistent treatise, knowledge of only isolated parts means that attempts to string them together will be necessarily “artificial” and totalizing – forced through a particular lens. Such a project will always be suspect, leading to an “inherent resistance to theory” in “literary” language theory. De Man claims that: [Rhetorical readings expose the structures and functions of language] but are an unreliable process of knowledge production . . . they are . . . consistently defective models of language’s impossibility to be a model language. They are, always in theory, the most elastic theoretical and dialectical model to end all models and they can rightly claim to contain within their own defective selves all the other defective models of reading avoidance, referential, semiological, grammatical, performative, logical or whatever. They are theory and not theory at the same time, the universal theory of the impossibility of theory. . . . Nothing can overcome this resistance to theory since theory is itself this resistance. (20) Even rhetorical readings which do indeed string parts together into an apparently consistent whole de Man claims are “totalizing (and potentially totalitarian)” (20) – an echo of Bacon’s  75  worries about presentations of science which are “fashioned and masked as if they were complete in all parts and finished” (Novum Organum LXXXVI). Arguably there are a host of good reasons why, in order to express insights implicitly that might be difficult or impossible to explain in more explicit consistent ways, sceptical language “theorists” would be prompted to employ stylistic indirection. Despite the imperfections of language and various insufficiencies of linguistically mediated human thought, as de Man observes, “to claim that this would be a sufficient reason not to envisage doing literary theory would be like rejecting anatomy because it has failed to cure mortality” (12). Not only do we not need x-ray vision, it is okay to wear glasses, or even contact lenses. Traces of this epistemologically inspired language scepticism can be found in Parmenides’ fragments and it is inscribed throughout the Platonic text. De Man argues that resistance to theory: can be read in the text of literary theory at all times, at whatever historical moment one wishes to select. One of the main achievements of the present theoretical trends is to have restored some awareness of this fact. Classical medieval and Renaissance literary theory is now often being read in a way that knows enough about what it is doing not to wish to call itself “modern.” (13) Recall that Derrida too sees “the question of the sign” as “an unexpectedly historical one,” and as if to corroborate certainty compulsions and theory-resistance tendencies observes that the “analogy between the structuralist obsession and the anxiety of language is not a chance one” (WD 3-4). Since the long-standing sceptical approach to language, which seems to have suffered from a certain resistance, is still insightful, this thesis underscores the affinities between the views of different thinkers from Plato to Derrida. A theory that embraces the historicity of the linguistic sign, that escapes determinate systematization, and that is observant rather than oblivious, might resist the totalizing-systematizing look of theory itself, might not want to call itself “theory” at all. Such an interest doubtless motivates Derrida’s expression “theoretical  76  matrix.”84 Inasmuch as its encounter with linguistic complexity deters the enterprise of “beholding” the contours of how the parts of language work as a self-consistent whole, sceptical (eyes-wide-open) language theory is itself theory resistant. Since it might not look like “theory,” the form of such a language “theory” begs to be investigated. In Part 5 we catch a glimpse of what a sceptical-theory-resistant-language theory might look like, and what its styles and genres can be.  84  In Of Grammatology Derrida outlines a “theoretical matrix” offered not as a “new method” but rather as an attempt to expose “the problems of critical reading” (lxxxix). He says grammatology is “a science of writing [that] runs the risk of never being established as such and with that name. Of never being able to define the unity of its project or its object. Of not being able either to write its discourse on method or to describe the limits of its field” (4).  77  PART 3 3.1  conn-texts & citation ~  whence the fascination for titles, which often function in a rather unsaturated context (Bennington 87)  Part 3 of this dissertation explores aspects of various contemporary language discourses that set the stage for Nietzsche, whose primary linguistic insights were overlooked as the positivist impulses of analytic philosophy gathered steam in the early twentieth century. In the mid-twentieth century the so called “philosophy of language” was consolidating its power, superceding sceptical-rhetorical theories. Whorf had been spurned as heretical; the rich rhetorical tradition had been relegated to an outpost postion and reduced to the study of political oratory; and the view that language was all about communication had been centralized. The Searle-Derrida dispute that resulted from the publication of a talk Derrida delivered to the 1971 Congrès international des Sociétés de philosophie de langue francaise in Montreal offers a window into the language-theory scene just as the sceptical-rhetorical theorists started demonstrating against analytic philosophy’s idealization of language. Although the very significant conference theme for this meeting of the French Language Philosophy Societies was “communication,” the rhetorical situation85 of the talk sparking the controversy – its context, its theme and its audience – seems to have got lost in the crossfire. Years after the conference, Derrida’s talk, “Signature, Event, Context” (SEC), was published. Then the fracas broke out with Searle’s pompous response “Reiterating the Differences: a Reply to Derrida.” But Searle’s view of language may not, at the time, have considered rhetorical situation especially important when intention-governed meaning was still current in the “philosophy of language.” Quite apart from the specifics of the audience, and apart from the talk’s inclusion of Austin on the same stage with other prestigious thinkers, Condillac and Husserl, Searle didn’t seem to notice the communication problematic in the published version of SEC. Derrida’s talk, though remembered for its discussion of context and citation, challenged the communication and code-metaphor assumptions underlying the philosophic texts of various 85  The term “rhetorical situation” was coined by Bitzer but equally addresses what prompts Burke’s “pentad” – act, scene, agent, agency, purpose.  78  thinkers, including Condillac, Husserl and Austin. However, perhaps because the talk was wrongly perceived as an attack on Austin, Searle didn’t recognize the importance of the communication theme. Or perhaps he just didn’t get the problem; perhaps he didn’t recognize the communication issue because he assumed that the purpose of language was communication, and thus it didn’t seem relevant. This failure evinces not only the recognition problems that haunt sceptical-language theory, but Sperber and Wilson’s findings that reader inferences determine relevance – not code-models of intention-governed meaning. Recall also Ellis’critique that (facile) code-communication models have dominated language theory; and Halliday’s announcement that Whorfian-type linguists had long since rejected communication as a model for how language works. Analytic philosophy’s idealizing tendency seems to have bundled logic, rule-governed grammar and communication-success together and come up with code-metaphor models of how language works; and whatever breakthroughs Austin had made, the model of intention-governed communication was still prevalent. In SEC Derrida sets out to disrupt “the authority of the code as a finite system of rules” and to question codecommunication assumptions that had directed investigations in early-twentieth-century “philosophy of language” (8). Querying the core of these assumptions, the opening line of SEC asks whether it is “certain that there corresponds to the word communication a unique, univocal concept, a concept that can be rigorously grasped, and transmitted: a communicable concept?” (Margins 309).86 Derrida’s opening line thus alludes to commonsense notions of communication: notions that fail to take into account the problematic nature of consciousness, intention, and the signifying structure of the sign. By this time, these were standard features of Derrida’s critique of writing previously explored in some detail in La Voix et le Phénomène that Searle doesn’t seem to know about. He was likely not an audience member of the French Language Philosophy Societies’ meeting in 1971. Derrida’s critique of the 1960s (like those of Burke, Whorf, and Kuhn) directed attention to problems in language – problems that in the end led to the demise of 86  Citations from “Signature, Event, Context” are from Weber’s Limited Inc. translation except where I have preferred Alan Bass’ translation in Margins of Philosophy.  79  the famous “linguistic turn [that] Wittgenstein initiated in the Tractatus” (Bergmann 63). This turn – once celebrated as the inauguration of analytic philosophy – was crucially already recanted by Wittgenstein himself in Philosophical Investigations. It was many years after Philosophical Investigations’ recantation, and a number of years after the 1971 talk itself when, in 1977, “Signature, Event, Context” was published in Glyph, and Searle replied, that the now infamous spat occurred. The arrogance of Searle’s “Reiterating the Differences: a Reply to Derrida” (which gratuitously defends Austin and also indexes the hegemony of analytic philosophy at the time) now smacks of irony. Speaking of communication gaps – not only did Searle not seem to recognize or grasp the implications of the communication question with which the talk opened, he seems to have missed Derrida’s expressed admiration for Austin altogether. Actually, Searle really didn’t seem to understand what Derrida was saying at all: he seems to have miss-taken what Derrida intended to communicate, specifically the critique of intention-governed, code-communication assumptions about language use. The question of being able to recognize sceptical-language theory thus presents itself in this infamous spat. On the one hand, Searle’s failure to recognize the important elements of Derrida’s critique may look like theory resistance, or seem to have been prompted by affective responses, or a bit of both. One might wonder whether Searle’s attachment to Austin blinded him to the Derridian critique the way the Royal Society’s science-project enthusiasm blinkered the uptake of Bacon’s views of language, or whether Searle just found the Derridian critique epistemologically worrying (likely both). On the other hand, quite apart from his apparent lack of effort, Searle might have an excuse for not understanding Derrida. He is clearly not Derrida’s addressee, not his target audience: a disciplinary-genre matter, an audience-design question that we discuss in Part 4’s discussion of style and that arises in Part 5 when we consider the readersubjectivity of Derrida’s text. In Part 4, when we examine the purported “plain style” of Searle’s disciplinary field, we see how the style of science and code-communication models get confused with conceptions of “ordinary language,” leading to dubious stylistic assumptions about clarity and simplicity. Most importantly the interest in citation and context in Derrida’s  80  talk is particularly germain to the scholarship on reported speech which dominates Part 3 making SEC not only an excellent segue-way but also a point of alignment between Derrida’s philosophising and various fields of discourse analysis. Human motives – such as those prompting Searle’s vituperation – re-appear on the stage in the analysis of reported speech. And the language thought ligature in contemporary neuroscience comes into view before we conclude Part 3 with Nietzsche’s proto-post-structuralist understanding of language. Searle’s recognition failure of the code-communication problem recalls the addressee of Sperber and Wilson’s relevance theory accustomed to code models which they importantly refute. Recall also their conclusion that language is required for the “information processing” aspect of linguistic interactions and not “communication” itself: here the beneficiaries of analytic philosophy are moving away from the core assumptions that triggered the linguistic turn to start with. Austin appears, as Derrida observes, “to consider speech acts only as acts of communication” (13) where sceptical-language theory does not think about language primarily through the lens of communication at all.87 But the problem here is not necessarily the use of communication to study how language works: there is nothing wrong with studying instances of linguistic communication in order to understand the operations of language as Sperber and Wilson’s insights indicate and as we shall see when we examine reported speech. The problem is unexamined assumptions about communication – assumptions unquestioningly linked to conceptions of rule-governed grammar that result in code metaphors. This confusion is, I suspect, largely a result of the “written language bias” that we examine shortly. Derrida’s talk, of course, pre-dates Sperber and Wilson’s bi-lateral idea of communication – an image of communication more like partnered improvisational dance than unidirectional-code transmission – and one that does not arrive on the scene for over a decade. Derrida’s critique is  87 Austin for example assumes a communicative model in the following: “When is a statement true? The temptation is to answer (at least if we confine ourselves to ‘straightforward’ statements): ‘When it corresponds to the facts.’ And as a piece of standard English this can hardly be wrong. Indeed, I must confess I do not really think it is wrong at all: the theory of truth is a series of truisms. Still, it can at least be misleading. If there is to be communication of the sort that we achieve by language at all, there must be a stock of symbols of some kind which a communicator (‘the speaker’) can produce ‘at will’ and which a communicatee (‘the audience’) can observe: these may be called the ‘words,’ though, of course, they need not be anything very like what we should normally call words—they might be signal flags, &c.” (Philosophical Papers, 121)  81  directed toward code’s naive notion of active speakers informing passive hearers – a quite different image of communication than Sperber and Wilson’s ostension and inference, alternations between pointing and probing, which correlates so closely with findings on reported speech. By 1971 – in the wake of La Voix et le Phénomène , De la Grammatologie and La Dissémination – SEC’s query about the meaning of the term communication, questioning whether the concept of communication has a structured centre, anchoring its meaning was standard Derridian fare. Presuppositions (that don’t seem to get scrutinized) about coded communication as a model for how language works prompt Derrida’s announcement that “one must first of all ask oneself whether or not the word or signifier ‘communication’ communicates a determinate content, an identifiable meaning, or a describable value” (1). What communication means is still a question, hearkening back to de Man’s complaints that people assume they know what language is. And we still need to ask whether or not we know what the term communication means, and what conception of it governs assumptions about language. Arguably in 2009 the sens «propre» of “communication” includes everything from social interaction (or sharing) to one way signal transmission between computers. Its habitual contemporary use over this vast range – as though it had a single determinate meaning, that we don’t need to think about – makes communication something of an empty term (a lot like democracy). We assume that we know what it means. But do we? It is not just that, like other signs, it lacks determinate, delimitable referents, but that it is so wanton that it can be all things to all people. It is what Burke calls an “ultimate term” – there is no discussion about what it means but it embraces a cacophony of perspectives and meanings (ROM 186-197). The indeterminate nature of the linguistic sign here obliquely (and playfully) alluded to  82  in Derrida’s reference to the “graphematic structure”88 of the locution is also of concern to Voloshinov, who not only claims that “multiplicity of meanings is the constitutive feature of word” (101), but also that, as a result, “the meaning of a word is determined entirely by its context” (79). Halliday too takes for granted Firth’s notions about the primacy of context and the indeterminacies of linguistic meaning.89 Not surprisingly, context is another star player in Derrida’s analysis of the linguistic-semiotic structure of the locution. Commenting again on the amorphous meaning of the term communication, Derrida observes that “the ambiguous field of the word ‘communication’ can be massively reduced by the limits of what is called a context” (2). Hence his talk, SEC, on the theme of communication, addressing the “graphematic” (linguistic-semiotic) structure of the locution itself is “concerned with the problem of context and with the question of determining exactly how writing [i.e. language] relates to context in general”(2). SEC’s examination of the relationship between the linguistic-semiotic structure of the locution and the idea of contextualization reveals the identifying role of citationality in the structure of the sign itself. This citation-focus together with other disseminations from Derrida’s talk (like, somewhat ironically, the written-language bias) make SEC a fortuitous topos from which to launch this section of this dissertation. By taking the spat back a step to the AustinDerrida encounter in SEC, Glendinning’s “Inheriting ‘Philosophy:’ The Case of Austin and Derrida Revisited” offers the opportunity to explore confusions surrounding “non-seriousness” and “ordinary language” in the scholarship of reported speech: confusions that so interestingly  88 “Graphematic” from grafhv or writing displaces “linguistic semiotic” for Derrida’s strategic purposes. The clause can be construed in this context as “the linguistic semiotic structure of the locution.” While admittedly, writing is the primary academic medium, and it has a positive status for its role in “tradition” in Husserl, and the “graphematic structure” is in writing, about writing, based on writing, and is nevertheless a critique that applies equally to the non-written linguistic sign – despite (and because of) all the factors that may have prompted Derrida to focus on writing rather the speech – I believe the gesture is largely strategic and opportunistic. In Of Grammatology Derrida claims that “everything that for at least some twenty centuries tended toward and finally succeeded in being gathered under the name of language is beginning to let itself be transferred to, or at least summarized under, the name of writing” (6). This move displaces prior conceptions of language with a concept of writing (writ large) I believe because a traditional conception of writing, as a derivative form of speech, offers fecund material for his critique. This contrivance foregrounds the opposition between the full plenitude of speech and its shadow (the secondary, inadequate writing), inviting the deconstruction of the dualistic assumptions on which western metaphysics is based. Presence is implicated in the code metaphor for communication which has resulted in so much confusion. His gesture of displacing “language” with “writing” provides a convenient site (or topos) for Derrida, although it has been frequently misinterpreted as the privileging of writing over speech. 89 “Firth consistently maintained that all study of language was a study of meaning; and meaning was a function in a context, where the ‘context’ was located both in various strata of language itself and in its situational and cultural environment” (OL 243).  83  bring Austin beneficiaries into harmony with Derrida. The fundamental requirement that a sign be recognizable and repeatable leads to Derrida’s determination that one of its defining features (whether visual or audio) is its iterability as a mark. “The possibility of disengagement and citational graft” Derrida claims, “belongs to the structure of every mark, spoken or written” (12). “Every sign, linguistic or nonlinguistic, spoken or written . . . can be cited, put between quotation marks” (12). Citationality, or reproducibility, here is an important identifying, structural feature of the sign. Co-incident to the importance Derrida attaches to citationality in the analysis of the sign, examinations of reported speech have become the focus of more empirically oriented research – perhaps prompted by Voloshinov’s view that reported speech is paradigmatically “speech within speech, utterance within utterance, and at the same time also speech about speech, utterance about utterance” (115). The research of the scholars working on reported speech examines a variety of genres: academic, literary, and various civic (i.e., legal and political) types. The findings of this work tread on the sacred ground of verbatim accounts: leaving open the question about what happens – in the legal system, the news media, academia – when the cherished notion of faithful-verbatim reporting loses credibility. The extreme import of faithfulness in reported speech – that is, verbatim, word-forword, exact duplications of utterances – in the high-stakes situations of news reporting, academic research and court-room testimony prompts Short, Semino and Wynne to develop a “context-sensitive account of the role of faithfulness in discourse presentation” (353). This futile gesture – perhaps interesting from a psychoanalytic perspective for hints of anxiety and even resistance – bears the imprint (once again) of Cratylian-code-metaphor models of language. Short, Semino and Wynne are responding to the verbatim-deconstruction heresies of scholars like Sternberg, Mayes, Tannen, and Clark and Gerrig, the latter of whom flatly state “the verbatim assumption is incorrect” (799). These scholars’ analyses of actual conversation, civic genres (like legal proceedings and news reports) and traditionally-defined literature successfully refute commonsense verbatim assumptions and determine that the concept of verbatim itself is a  84  fallacy. They challenge not only faith that the Cratylian-code-metaphor, word-for-word reproduction of an utterance conveys the original speaker’s intentions, but also whether verbatim is anything but an idealizing concept when citation in fact involves “tearing a piece of discourse from its original habitat and recontextualizing it within a new network of relations [that] cannot but interfere with its effect” as Sternberg observes (108). Derrida too argues that the fact that every sign can be cited means that “it can break with every given context, engendering an infinity of new contexts in a manner which is absolutely illimitable” (12). This is hardly a surprise. Being “quoted out of context” is a well polished phrase – a common complaint – widely recognized as a way of misconstruing someone’s words. Margerie Garber’s “Quotation Marks” observes the obvious when she notes “that every quotation is a quotation out of context” (669). Similarly Geoffrey Bennington states: “One quotes by definition out of context” (85). Since notions of context have taken up the burden of the codemetaphor – as Short, Semino and Wynne’s interest in a “context-sensitive account of the role of faithfulness” reveals – context now performs the labour of determining meaning. But how determinate is it? The importance of context does not alter the fact that it is a slippery and recursive creature itself. Bennington’s Derridabase90 states, first of all, that the idea of the sign is inoperable without context; and secondly, context is recursive, contexts exist within contexts – like the sign, context is subject to the determination and delimitation problem of recursion mise en abîme: We [Bennington and Derrida] say “quote in another context” rather than “quote out of context” to mark the fact that there are always contexts: the logic of the trace makes the idea of a sign or statement outside any context unthinkable, while making possible an exploitation of very open contexts (whence the fascination for titles, which often function in a rather unsaturated context). (emphasis added 87) Every element of the context is itself a text with its context. There are only contexts, and one cannot proceed to make the usual text/context distinction unless one has already taken the text in itself, out of “its” context, before 90 Bennington’s text presents various currents in Derrida’s thinking (“if not the totality of J.D.’s thought, then at least the general system of that thought”) in Bennington’s Oxford-schooled language playfully styled as his “pedagogical and logical norms” (1).  85  demanding that it be placed back in that context. And if one accepts what has already been said about consciousness and intention, which rules out the possibility of ever deciding rigorously what a text “means,” then we must accept that every demand to put things back in context is already interested and cannot be neutral. (emphasis added 90) Thus Derrida sets out in “Signature, Event, Context” – the talk delivered to philosophers on the theme of communication – not only to disrupt “the authority of the code as a finite system of rules” but also to disavow abuses of the concept of “context as the protocol of code” (8). Derrida’s suggestion here that “context” is being exploited as if it were code explains how scholars who know very well that citations are always taken out of context can nevertheless cling to the importance of verbatim faithfulness – that the exact words will duplicate the intentions of the (original) speaker – thus to code models of communication. One of the larger contexts of Derrida’s paper then might be described as a conversation between analytic and “continental”91 philosophy. Searle’s miss-take of Derrida’s critique – taking the critique of the assumptions to which Austin is heir as dismissive of Austin himself – is festooned with ironies beyond Searle’s inability to understand Derrida’s intentions.92 Searle’s attack on Derrida and the subsequent hullabaloo and misunderstandings make it ironic that it is Austin who figures in “Signature, Event, Context” (SEC). It is ironic because Austin’s work also undoes code-conceptions of language (or at any rate, it starts this work) and it subsequently motivates linguistic pragmatics researchers who end up reaching conclusions a lot like Derrida’s.93 Derrida’s texts are always parasitic, always opportunistic (that is, always emerging from carefully selected topoi), always writing about writing: thus Austin’s text is very useful to 91  This is the name analytic philosophy seems to have given to those it wanted to exclude, that is, the name given to those who did not follow the “linguistic turn” but instead continued to challenge historical philosophic traditions. See Grondin’s “Continental or Hermeneutic Philosophy.” 92  Austin’s questioning of his own tradition’s assumptions – much as he works within them – is what attracted Derrida to his work to begin with. Derrida says that Austin’s “critique of linguisticism and of the authority of the code, a critique based on an analysis of language,” was what “most interested and convinced [him] in Austin’s undertaking”(emphasis added SEC 19). 93 The assumption that language use is descriptive goes with code conceptions. The performative dimension of language is precisely an effect, a force of language unwarranted by its descriptive-referential meaning. Observe this undermining trajectory on the opening page of How to do things with words: “It was for too long the assumption of philosophers that the business of a ‘statement’ can only be to ‘describe’ some state of affairs, or to ‘state some fact’, which it must do either truly or falsely. Grammarians, indeed, have regularly pointed out that not all ‘sentences’ are (used in making) statements: there are, traditionally, besides (grammarians’) statements, also questions and exclamations, and sentences expressing commands or wishes or concessions. And doubtless philosophers have not intended to deny this, despite some loose use of ‘sentence’ for ‘statement’” (1).  86  the SEC critique-of-code-communication-model enterprise. This is not only (but significantly) because Derrida admires Austin’s active, flexible, rigorous thinking (depicted as “patient, open, aporetical, in constant transformation . . . fruitful in the acknowledgement of its impasses”), and not only because Austin participates in a “way of going on” in philosophy94 analyzing language as code-communication, but also because Austin’s insights into “performative” verbs offer Derrida a fruitful opportunity – a topos or place – to explore, among other things, the interdependence between the iterative nature of the sign as a mark (or citationality) and context (14). Apart from all the ironies, Austin’s work is itself part of the context of Derrida’s talk at (at least) two levels: as both a site of analysis – the “place” giving rise to Derrida’s query – and as contributor to the larger scholarly conversation about language into which Derrida speaks. 3.2  legacies: code-communication and the written language bias ~  Admittedly the speakers gathered together in this thesis might not happily share such close quarters, given the sometimes considerable differences between them. That Voloshinov’s and Halliday’s thinking lines up so well with that of Derrida95 might be less surprising than that the “mainstream” linguistics-related work on communication and reported speech finds such close affiliation with Derrida’s. Scholarly affiliations and lineages, characterized as “inheritance” in Glendinning’s “Inheriting ‘Philosophy:’ The Case of Austin and Derrida Revisited” thematizes different “ways of going on” in philosophy – not broadly with the terms “analytic” and “continental,” but in more nuanced terms of style, clarity, obscurity and ordinariness. Much “mainstream” linguistics seems heir to Austin’s legacy, Austin himself being heir to code-communication-model assumptions about the truth status of propositions – precisely the assumptions Derrida questions in SEC. However much Austin’s own work actually undermines the code-communication orientation to which it is heir – assumptions treating codecommunication as the model for how language works persist in his thinking. What Glendinning 94  Glendinning uses this phrase to temper gross distinctions between these two “traditions” in philosophy, in order to defuse the hostilities (308). Derrida refers typically to someone’s “style” of philosophy or thinking. 95 Holquist observes that Bakhtin’s “concept of heteroglossia is a happy redaction of the conditions otherwise so gloomily charted by Derrida’s epigones as ‘differance’” (“Introduction” xxxii). Whether Bakhtin is Voloshinov or simply a very close colleague sharing many of Bakhtin’s views is not relevant here. The proximity of Derrida’s and Voloshinov’s understandings of language is clear. Similarly, Halliday’s Whorf connection makes him an easy bedfellow with Derridaaffiliates.  87  calls “truth-evaluable content” seems to act as a reference-post not only in Austin’s work but in subsequent generations of scholars following him (i.e., philosophers like Grice, psycholinguists like Clark and Gerrig, sociologists of language like Goffman, linguistic-pragmatics scholars like Sperber and Wilson). Although I doubt the presumed success of oral communication (an intuition that may never have benefited from transcription verification), I suspect the recidivism of code, that Ellis observes, stems from just such unexamined conceptions of communicative success: and that these unexamined conceptions are then incautiously affiliated with the logical-truth-status-ofpropositions.96 Unexamined intuitions about communication working so well seem to be superimposed on beliefs about the role grammar plays; communication is perceived through the lens of rule-governed grammar (as the logic of language); and then without pausing for breath rule-governed grammar is perceived back through the lens of communication. Communication works, grammar is a kind of logic, therefore communication must depend on logical deduction. Wittgenstein’s insight that the “requirement” of “the crystalline purity of logic” was like a “pair of glasses” through which to see – a pair of logic goggles – makes it easy to see how these communication intuitions might have prompted logic-bespectacled philosophers to fantasize conceptions of rule-governed grammar, truth-evaluable content and propositions in order to account for assumptions about communication success. Voloshinov’s complaints that such rule-governed grammars stem from analysis of written texts direct us to some speech and writing confusions embedded in this code-logic approach. The model of communication in which meaning determination results from the truth status of propositional content depends on sentences, artifacts of writing not speech. When we look at transcripts of actual speech (as we will with Holt and Tannen) it becomes clear that people only occasionally use whole sentences when they speak to one another. If philosophers  96 It seems unlikely that writing – famous for communication failures and producing a plurality of interpretations – would have prompted notions of communication success. As Socrates notoriously observes “if you question [written texts], wishing to know about their saying, they always say only one and the same thing. And every word, when once it is written, is bandied about, alike among those who understand and those who have no interest in it, and it knows not to whom to speak or not to speak” (Loeb Phaedrus 275e).  88  supposedly focussed on “ordinary language,” like speech-acts theorists, partnered intuitions about oral-communication success with written sentences – as though speech saw an image of writing when it looked in the mirror – then we can see how Per Linell’s The Written Language Bias in Linguistics gets its start.97 Writing within the field of linguistic pragmatics, Linell and Marková fault speech-act theory precisely for its “written language bias.”98 They complain that the “written language bias in linguistics” works on sentences as autonomous units and made up examples (instead of actual discourse),99 and that it “accepts the idea that the task of semantics or pragmatics is to map onto each possible expression a unique semantic interpretation” (“a determinate content,” as Derrida says, with “a concept that is unique, univocal, [and] rigorously controllable”) (174).100 Linell and Marková claim that speech act theory is more of “a supplement” to, than a replacement for, “traditional sentence-based semantics,” which is concerned with “truth conditions and propositional content” (174). Not only was speech presumed to act grammatically like writing but, as Linell and Marková remark, the supposedly oral sentences analyzed were often inventions of the philosophers studying them – despite the radical changes in context between most written and spoken utterances. Successful communications – whether written or spoken – do not depend on truth-conditional rules as the code model imagines. It is much more likely that they depend on relevance “rules” – if you can call them rules when they look more like dances or games. Derrida was doubtless justified in detecting a myriad of unexamined and muddled  97  Per Linell’s introduction opens: “In this book the modern linguist's view on language is discussed from a rather unusual point of view. It is argued that our conception of language is deeply influenced by a long tradition of analyzing only written language, and that modern linguistic theory, including psycholinguistics and sociolinguistics, approaches the structures and mechanisms of spoken language with a conceptual apparatus, which - upon closer scrutiny - turns out to be more apt for written language in surprisingly many and fundamental aspects. I will refer to this situation as the written language bias in linguistics.” 98 While it is clear that Austin works with code-communication and truth-conditional reference posts, this critique does not appear to be directed towards his the very initial and tentative work How to Do Things With Words. Rather Linell and Marková’s critique is directed to “speech act theory as developed by John Searle” (173). 99  Interestingly Linell and Marková criticize “the written language bias” in “Chomskyan generative grammar, Grice’s theory of conversational implicature, Sperber and Wilson’s theory of relevance, and Labov and Fanshel’s model for analyzing real talk” (175). 100  For example, Clark employs the following example of supposedly oral conversation: “Ann, to Charles, in front of Barabara: Charles, I insist that Barbara tell you who we met at the museum today.” (Arenas 225). Regardless of what Clark’s point is here, the example reinforces the myth that people normally speak in sentences like this, Ann is more likely to have said “ask Barbara who we met at the museum today.”  89  assumptions surrounding the term “communication” in the speech-act-theory discourses. 3.3  legacies: performatives, seriousness & ordinary language ~  Nevertheless Austin’s ideas about the way that actions are performed by language use (language acting rather than referentially describing) represented by his term performative did rejuvenate ways of thinking about how language works within analytic philosophy and its “mainstream” linguistics beneficiaries. Glendinning’s “Inheriting Philosophy,” which adjudicates “the case of Austin and Derrida,” outlines the arguments in Austin’s work on “performative communication” that exclude “non-serious” speech events from “serious” ones – a problematic and unfortunately named distinction. Glendinning is a particularly useful guide here because, as an analytically trained philosopher, he shares Austin’s heritage, but he also reads Derrida. It is easy to see how Austin (the alert thinker that attracts Derrida’s attention) got stuck in problematic code-communication assumptions when Glendinning traces the serious/non-serious distinction back to Frege. He shows that Austin’s use of this nebulous and intention-governed distinction is a Frege legacy – a bit of discourse, available as a ready-tohand assumption whose implications did not require assessing. Since we must start with assumptions – a double bind Plato, Nietzsche and Derrida have all wanted to foreground – being an intellectual beneficiary is doubtless always complicated.101 The vicissitudes of heritage are manifest when we find this distinction from Frege reappearing in Austin legatees Clark and Gerrig who (ironically) link it to citation in order to argue, as Derrida does, for the ordinariness of the “non-serious” in the ubiquity of quotation. When Austin writes How to Do Things with Words serious/non-serious seems to mean something like logical/unlogical, literal/non-literal, real/unreal. While I doubt the usefulness of such distinctions, (especially the literal/non-literal which depends on a Cratylian rather than a tropological conception of language), Glendinning shows that Austin himself was suspicious of such dualistic metaphysical assumptions. Citing Sense and Sensibilia, he observes that (with regard to perception) Austin claims to “‘abandon old habits of Gleichschaltung, the deeply 101  Glendinning states that “the problematic of inheritance is itself a central and inescapable issue” of the topic he has chosen to write on.  90  ingrained worship of tidy-looking dichotomies’” (311). Gleichschaltung, according to Glendinning, means bringing “something into line” making “something conform to a certain standard by force”; he also elaborates on the term’s Nazi political affiliations (illustrating the way that concerns for “justice” are implicated in the Derridian critique) (312). The term Gleichschaltung and Austin’s phrase “tidy-looking dichotomies” will prove useful to (later) considerations of Aristotelian stylistics; however for current purposes Glendinning cites Austin’s non-serious exceptions as being guilty of the same Gleichschaltung and tidy-looking dichotomies that Austin criticizes. But perhaps Austin just reached for this terminology (criteria) without pause (as Bolinger and Becker have suggested that we reach for a phrase, “a remembered prior text”) as a convenient category to narrow down his field of study. Glendinning observes that Austin’s exclusion of “non-serious” speech acts from the analysis of performative verbs calls into question the concept of ordinary language. He says that “Austin’s concept of ordinary is, as Derrida puts it, ‘marked’ by this exclusion” (326). Summarizing Derrida’s objection that Austin is “insufficiently sensitive” to the possible risk of things “going wrong” in speech acts (irony or sarcasm counted among the non-serious “infelicities,” for example) Glendinning writes: Instead of pursuing an investigation of the structure of locutionary acts which shows why this ‘risk’, qua possibility, is essential to its being such an act, [Austin’s] procedure passes over the general logic of locutions and positions it, qua eventuality, as something that transgresses the ‘ordinary circumstances’ of language use – something that can, and therefore should, be pared away or excluded from an analysis of the ‘normal use’ of words in ‘ordinary circumstances.’ (327) It is the very idea of excluding “ordinary” types from analysis of “the ordinary” that is at issue. It is not that Austin’s insights about the behaviour of certain verbs were not useful or interesting; it is that “ordinary” and “normal uses” of language are sidelined by assumptions about communication that exclude more fundamental insights about the semiotic-linguistic structure of the locution. Of course for Derrida the impossibility of determining and delimiting intentions – on which so much of the non-serious category depends (along with its Cratylian  91  implications) – makes the infelicities problematic to start with.102 But Glendinning focuses on the fact that Derrida does not believe that the exceptions can be ruled out in the investigation of the ordinary. He continues: Unless analytical philosophy is to be characterised by a lack of concern for the structure of ‘ordinary language’ this [ruling out of exceptions] cannot be so. For the reason why Derrida is convinced that consideration of ‘the realm of the “nonserious”’ cannot be put off until later can only be understood in terms of such a concern. Specifically, his claim is that because it is internal to the structure of ‘the realm of the “serious”’ (the realm which Austin’s analysis aims, here and now, to consider) that it possesses features held by Austin to define the excluded realm . . .” (327). As heir to problematic-communication assumptions, when Austin reaches for the ready-to-hand label “non-serious,” he fails to perceive that non-serious uses are internal to the structure of serious uses. Perhaps it isn’t just a coincidence that Derrida’s talk – claiming that citationality is structurally necessary identifying feature of the sign and finding that so-called non-serious uses of the locution are internal to the structure of serious ones – jives so well with the findings of scholars studying reported speech, the scholars studying citation. Clark and Gerrig’s analysis of reported speech determines that quotations, what they call non-serious “demonstrations, are performed as parts of serious activities” (766).103 Arguing that citations are “demonstrations,” a type of “showing” likened to Wierzbicka’s description of quotations as “imaginary speech performances,” and (following Goffman) that “demonstrations” are “non-serious” (meaning “not literal”) they conclude that reported speech constitutes a substantial component of language use, i.e., these “non-serious” performances are ordinary, normal usage (801, 766). The footprints of the dubious “serious/non-serious” distinction lead from Frege to Clark and Gerrig. Clark and Gerrig attempt to distinguish non-serious demonstrations, like quotation (which they claim depicts reality), from serious, descriptive uses of language. I believe that these categories are unsustainable: the terms serious and non-serious bring with them their own  102  Who can determine and delimit even their own intentions? As Socrates says: “our soul at any one moment teems with countless . . . contradictions” (Republic 603d Loeb) 103  There is an irony in this identification of citation as non-serious, since, as Derrida points out in SEC, the original list of performative verbs were citation par excellence.  92  histories of association and affiliations – especially code conceptions of communication and literalness. Though Clark and Gerrig attempt to provide descriptive classification (which from their perspective is “serious,” non-demonstrative language use) Foucault might remind them that description is inevitably prescriptive – prescription being a by-product of descriptive classification. Halliday observes language contrues reality – that is, constructs reality – and enacts it at the same time. As Austin’s own analysis begins to reveal (when he extends illocutionary force to all speech acts and adds describing to the list of “performative” verbs) and as any language usage manual clearly illustrates, descriptions act in the world, producing what they seem to describe – both reifying (constructing) and prescribing (enacting). Austin’s inquiry into the actions of a very limited number of performative verbs lodges a speech-as-action wedge into description-referential code-communication models of the structure of the locution dominating analytic philosophy. This wedge opens up a treasure trove of linguistic insights, within analytic philosophy, about language’s attributes. These pragmatic insights are congruent with ancient and contemporary sceptical views of language, and lead to this thesis’ use of the term performativity. Hoping that a fuzzy picture is exactly what we need, language’s performativity gathers together diverse effects of language in which it seems to act: figurative effects, labelling/categorizing/reifying effects, (interpersonal) enacting effects, and so on. For example, the Foucauldian productive-creating-prescribing-type acting in the world – that is, language that both produces what it seems to describe (i.e., mental or physical disorders), and that then disciplines by pre-and proscribing – illustrates two ways that language performs actions. As it turns out, citation, which constitutes a much larger portion of every day language use than the initially-identified performative verbs, is paradigmatic of language’s performative disposition. 3.4  intersections: language theory and reported speech ~  Since Wittgenstein’s turn to “ordinary” or “everyday” language is a turn to how we actually use language – in everyday conversation, in news reports, in academic writing – arguably scholars of reported speech really did turn back to “ordinary language.” Among the  93  curiosities arising from the scholarship on reported speech, one that speaks to its “ordinariness,” is that results from analysis of (what would be considered) “literary” texts line up with analysis of a wide range of everyday language genres like transcripts of actual conversational discourse, courtroom testimony, legal documents, and journalism. Tannen’s analysis of conversational talk leads to her observation that the strategies used in everyday talk are the same ones that she had previously considered “quintessentially literary” (30). This overlap between the strategies of conversational talk and literary writings make spontaneous talk an exemplary site for analyzing the mechanics of language use. Meyers too comments on the link between “work on spoken language” and “literary stylistics” in studies of reported speech: prompting the observation that “everyday talk is in its way very artful” (Reported 378). Whether this is life imitating art, or art imitating life, (or perhaps the art of linguistic life), what seems clear is that reporting other’s speech is so fundamental, so consistent and persistent in language use, that it simply occurs where language is used. As Derrida has reminded us it wouldn’t be linguistic semiotic phenomena if it were not iterable, repeatable, re-sayable. Citation really is ordinary. Analysis of reported speech produces unexpected and counterintuitive results in which descriptive reference once again leaves the scene: citation seems to do other types of work more in line with Austin’s actions. Recall that de Man defines “literariness” as “the use of language that foregrounds the rhetorical over the grammatical and the logical function” (14). Examination reveals that reported speech is a rhetorical device employed for a variety of strategic purposes (often functioning at multiple layers at the same time). Thus Fludernik writes that “so-called verbatim representations of utterances do of course, exist,” however, counter to expectation socalled verbatim representations are not “the most mimetic, natural, unmarked form of speech report,” rather they are “the most highly artificial, marked and formulaic type of reported discourse” (2). Direct reported speech – that seems the most “natural,” “transparent,” unmarked, “faithful” type of report – is the most contrived. The same kind of findings lead Holt to argue that “direct reported speech is a complex device characterized by contradiction” (427). Oddly, some kinds of language use seem designed to deceive; what seems most mimetic is the most  94  artificial. Whether these contradictory and counter-to-expectation characteristics of direct report indexes intuitions about language’s descriptive-referential nature (that may be schooled and historical) or whether they are just another manifestation of linguistic complexity is not clear. The convergence of literary, philosophical and linguistic scholars whose work is coalescing here around citation covers a broad range. Derrida brings a philosophic approach to the question of linguistic mediation which is epistemological and structural; and Voloshinov (whose advocacy of studying reported speech prompts much of the pragmatics work presented here) offers a Bakhtin-Circle-“literary” type theorizing, which celebrates reported speech’s active reception of another’s words together with one’s own evaluation of them. On the other hand, most scholars studying reported speech empirically analyze instances of actual reporting in texts often from corpus studies. For example, Deborah Tannen’s “discourse analysis,” gathers together research mainly in anthropological linguistics supplemented with socio- and psycholinguistics and her own background in literature and literary theory (5). This work on conversational discourse observes the ubiquity of repetition (not just in the iterative nature of the sign) but more broadly in the human mimetic impulse (what Nietzsche calls the drive to imitate). Mimesis is also a fundamental analytic concept in Sternberg’s seminal literarytheoretical, discourse-studies type exploration. Sternberg’s analysis of the five attributes of reported speech (affect, detail, realism, distinctiveness, reproductivity) in texts (ranging from Plato to Dick Francis) prompts him to assert that all reported discourse, regardless of form (or directness), is a mimesis of discourse (107). His much cited “Proteus in Quotation-Land” underwrites Fludernik’s exhaustive book-length study of forms of reported speech (most especially free indirect discourse). Her approach in The Fictions of Language and the Languages of Fiction situates itself squarely “between a linguistic and a literary analysis of narrative prose,” much like other “discourse analysts” cited here including Meyers, Calsamiglia and Ferrero, and Holt. (12). Reviewing samples of actual recorded conversation makes the sociality of spoken communication immediately clear. Holt’s transcription protocol, which captures non-verbal  95  vocalizations and linguistic dialect, implicitly indexes just how much communication depends on body language and non-verbal gestures: a reminder of what is edited out when supposedly “verbatim” accounts (such as news reports) normalize dialectical usage and accent, and omit various expressive features like pauses and tone of voice. For example, Holt’s transcript below of a daughter narrating an incident to her mother shows the importance of the mother’s active signs of interest, comprehension, and concurrence, evinced by various verbal and non-verbal vocal gestures. Holt is especially interested in the implicit evaluation in the direct report of Mrs. Field’s utterance “oh, I thought Lesley had given me these” (ll 6-7). While this statement seems banal enough out of context, within the context it acquires sharp edges, expressing Lesley’s clear distaste for something about the reported speaker. Lesley is reporting looking for some books. 1 2 3 4 5 6 7 8 9 10 11 12 13 14  and em I couldn’t hfind them ‘n then hsuddenly we realized th’t hI’d mi- I: lent’n to Missiz Field (0.2) hyears ago. Mum: Oh:. Lesley: •hhh An’ hMark a:sked her f-back ifor th’m this g evening ‘n she said •hh hOh: I thought Lesley g had given me the:se (0.3) Mum: iOh:.uh hhe -:-:h Lesley: .NO h: such lu:ck ihh hhuh Mum: .Mm-i:::hHm:. (0.6) Lesley: When you isee ‘er you’d better iwa:rn her. (Holt 444) Lesley:  It is easy to see that Lesley’s report of Mrs. Field’s utterance is a negative assessment of Mr