Open Collections

When if or when specify modals∗Jos Tellings, Utrecht University, j.tellings@gmail.comWCCFL 38, UBC, 07 Mar 20201 Introduction: namely as a speci€cational adverb• Namely is a speci€cational adverb that introduces an answer to an implicitquestion that is raised by its antecedent:(1) Mary climbed a famous mountain, namely Mt. Blanc.implicit Q: which mountain? answer• Dutch namelijk and German na¨mlich, in addition to the speci€cational use,also have an explanative reading, that I will not be concerned with today(see Onea and Volodina 2011 for more on the explanative reading).1• Typical antecedents of namely are inde€nites (1), certain de€nite descrip-tions (2), or wh-constituents (3).(2) Fred scaled the tallest building in the world, namely Burj Khalifa.(AnderBois and Jacobson 2018: 392)∗I thank the audience at the ILLC in Amsterdam for their valuable comments and feedback, whenI presented this work at the Meaning, Logic, & Cognition (MLC) seminar. ‘is research is part of theTime in Translation project, funded by NWO grant 360-80-070, which is gratefully acknowledged.Project website: h‹p://time-in-translation.hum.uu.nl.1A Dutch example of explanative namelijk is the following:(i) [Dutch]LindaLindaisisblij.happyZeshehee‰hasnamelijknamelyhaarherwerkworkaf.€nished‘Linda is happy, because she €nished her work’‘e explanative use of namelijk/na¨mlich can also be distinguished from the speci€cational use, be-cause the two have di‚erent syntactic properties. See (33) in the Appendix for another example, andOnea and Volodina (2011) and Onea (2016) for further discussion.(3) I ate what Mary cooked, namely ratatouille. (AB&J: 392)• In this talk I will consider cases in which namely takes a modal expressionas antecedent, and an if - or when-clause as complement:(4) Working as a €lmmaker can be taxing, namely if you’re required toget sleek product shots. [iWeb corpus]‘ese data are worth analyzing for two reasons:1. Cases like (4) are not covered by proposed generalizations about thedistribution of namely. ‘ey thus tell us something about the behaviorof the speci€cational adverb namely, as well as the question-raisingpotential of modal operators.2. Most commonly if -clauses are studied as adjuncts, i.e. in conditionalu‹erances. In (4), the if -clause takes a di‚erent position, as well asa di‚erent semantic role. Both constructions have an interaction be-tween a modal expression and an if -clause, but di‚erently so: modalrestriction and modal speci€cation are distinct.2 ‡e modal namely construction2.1 Types of modals• I will refer to the construction in (4) as themodal namely construction. ‘econstruction also exists in other languages (see Appendix), but I will focuson English in the talk. Below are some more examples, all drawn from theiWeb corpus:(5) a. Working as a €lmmaker can be taxing, namely if you’re requiredto get sleek product shots.b. IIN may ask for personal data, namely when you: request a cat-alog or magazine, order books, take our quiz, subscribe to our1Jos Tellings When if or when specify modalsnewsle‹ers, or request customer service.c. ‘e resultant equilibrium distribution can be di‚erent from aMaxwell distribution, namely if the situation is not isotropic.d. It means that in a world that is always mediated by di‚erent me-dia, one can still be free, namely if one knows how to operate themedia that dominate one’s world;e. Mixing colors on the computer can be challenging, namely whenyou are trying to translate speci€c colors to work in a four-colorprinting process• Note that namely can also take an if- or when-clause as complement with anexplicit nominal antecedent such as ‘case’ (6) or ‘circumstance’ (7):(6) However, there might be certain cases where you want to use it,namely if the opponent has way too many demolishers (4 or more)[. . . ](7) A breach of the peace may take place on private premises but only inde€ned circumstances, namely if a member or members of the publicare likely to be disturbed.• Not all modal expressions are a good antecedent for namely. First, universalforce modals and veridical verbs are disallowed:(8) a. Working as a €lmmaker must be taxing, #namely if you’re re-quired to get sleek product shots.b. Working as a €lmmaker is taxing, #namely if you’re required toget sleek product shots.‘is is the same pa‹ern we see in other constructions involving implicitquestions, such as sluicing ((9); from Chung et al. 1995: 254), and nominalnamely, (10):(9) a. Joan ate dinner with several students in her class, and we’re allwondering (with) who.b. *Joan ate dinner with every student in her class, and we’re allwondering (with) who.(10) a. Joan met a student from her class, namely Peter.b. ?Joan met every student from her class, namely Peter, Linda,Mary, . . .• Moreover, not all existential force modals are good with namely. Considerthe example below:(11) You can get a refund, “namely if there was a production fault.Intuitively, if one says “you can get a refund”, it doesn’t mean that thereexist circumstances in which you get a refund, but rather that in your cir-cumstances, you get a refund.• Namely is also degraded with epistemic modals:(12) (For all I know) Linda may be in her oce, “namely if she has ameeting with her student.In order to understand these restrictions, we have to understand the type ofmodal expression used in (5), see §2.2.2.2 …anti€cational modals• Portner (2009) de€nes the (somewhat understudied) class of quanti€ca-tional modals.A quanti€cational modal “incorporates the semantics of an adverb of quan-ti€cation together with some sort of additional, more properly ‘modal,’ mean-ing” (p. 213). Below are Portner’s examples (p. 214):2Jos Tellings When if or when specify modals(13) a. A dog sometimes bites. ↔ A dog can bite.b. A dog always bites. ↔ A dog will bite.adverb of q. q. modal• I want to argue for the following descriptive generalization:(14) GeneralizationOnly existential quanti€cational modals can serve as modalantecedents of namely.• In the examples of modal namely in (5), each modal can be paraphrased withan adverb of quanti€cation: “working as a €lmmaker is sometimes taxing”,“IIN sometimes asks for personal data”, etc.• In the other direction, we note that in (11), “you can get a refund” does notmean “you sometimes get a refund”, so can does not have a quanti€cationalmodal reading here. ‘e same holds for the epistemic modal in (12): “Lindamay be in her oce” 6= “Linda is sometimes in her oce”.• As further support for the generalization in (14), we €nd that overt temporalquanti€ers such as sometimes or occasionally combine with namely (exam-ples from iWeb corpus):(15) a. Fear also hardens sometimes, namely when it is not great.b. It has been pointed out that in British English at least, the “of”in “could of” etc. is sometimes clearly audible, namely when theword is stressed.c. She only occasionally wears that engagement ring from Je‚,namely when she wants extra a‹ention.• Overview of possible antecedents of namely if / namely when:∃ quanti€cational modalsNPs with Ns such as ‘case’, ‘circumstance’∃ temporal quanti€ers(‘sometimes’, . . . )namely if/when . . .3 ‡eories about namelyDi‚erent theories have been proposed about the precise nature of the implicitquestion that namely answers. I’ll discuss below AnderBois and Jacobson (2018)(henceforth AB&J), and Onea (2016).3.1 AnderBois and Jacobson (2018)• AB&J claim that the antecedent of namely needs to introduce a discoursereferent. ‘e implicit question is then a speci€cational question about theidentity of that discourse referent.• ‘is is reminiscent of sluicing, which has been analyzed as requiring an an-tecedent that makes an inquisitive contribution (AnderBois 2014). However,AB&J point out a number of di‚erences between both constructions. One ofthem is the availability of ‘sprouting’, which is €ne with sluicing (16a), butnot with namely (16b) (p. 399):(16) a. Juan celebrated his graduation, but I don’t know where.b. * Juan celebrated his graduation, namely on the beach.According to AB&J, these data show that sluicing and namely-constructionsinvolve di‚erent sorts of implicit questions.• ‘ey propose the following generalization:3Jos Tellings When if or when specify modals(17) Namely generalization (AB&J: 395)“namely is licensed i‚ (i) there is material in the preceding discoursewhich supports a discourse referent, and (ii) the fragment serves tofurther specify that discourse referent.”3.2 Onea (2016)• Onea (2016) instead argues that the implicit question that namely answers isdirectly derivable from its antecedent. ‘is is part of a general theory of howu‹erances have the potential to raise implicit questions into the discourse.2‘e notion of a ‘standard potential question’ (SPQ) is rather weak:(18) Standard potential question (Onea 2016: 120)A SPQ q licensed by some u‹erance u in some context c is such that:(i) c ∩ JuK |= p and c 6|= p, where p = info(H(q))↓;(ii) and there is no p′ ∈ H(q), such that c ∩ JuK |= p′.(whereH(q) denotes the union of the highlighted alternatives of q;see Onea 2016: §3.3 and Roelofsen and Van Gool 2010)• Examples of some SPQs raised by the sentence ‘Mary danced’:(19) Mary danced. (Onea 2016: 125)SPQ: p = info(H(q))↓:Who is Mary? that Mary is somebodyWhen did she dance? that Mary danced at some past timeWhere did she dance? that Mary danced at some placeWhy did she dance? that Mary danced for some reason...2In the recent framework of dynamic inquisitive semantics (Dotlacˇil and Roelofsen 2019), theframework is extended with discourse referents and issues about their identity (‘?x’). ‘is o‚ersthe potential to integrate insights from AB&J about discourse referents and their correspondingspeci€cational questions, with the inquisitive approach from Onea.• Namely is sensitive to a more restricted type of potential question, namelya primary potential question (PPQ).(20) Primary potential question (Onea 2016: 133)A potential question q licensed by some u‹erance u in some contextc is a primary potential question (PPQ) licensed by u in c, i‚ theset of highlighted alternatives in q is compositionally derived/madesalient by u.For example, the sentence ‘Mary danced’ in (19) does not license any PPQs.• ‘e notion of PPQ is intended to explain familiar contrasts such as the fol-lowing (Onea 2016: 134):(21) a. Peter was with someone, namely with John.PPQ: Who was Peter with?b. *Peter was not alone, namely with John.no PPQ (only SPQs)• Onea formulates the following generalization:(22) Onea’s generalization about namely (Onea 2016: 43)“In English, namely only addresses primary potential questions (PPQs)which can be derived from the main u‹erance by replacing the an-chor [=antecedent] with a wh-word.”3.3 Back to the modal data• ‘e modal data are a counterexample to Onea’s generalization: the modalexpressions don’t license a PPQ. ‘ere is no wh-word that can replace themodal expression.Maintaining Onea’s account would require that the modal raise a PPQ of thesort “Under what conditions/When is working as a €lmmaker taxing?” for(8a), but this does not follow from the de€nition in (20).4Jos Tellings When if or when specify modals• ‘e implications of the modal data for AB&J’s account are a li‹le morecomplicated. Does a modal expression introduce a discourse referent? Inmany analyses of modal subordination this has indeed been proposed (seee.g. Stone 1999; Brasoveanu 2010 a.o.).(23) Au1 wolf mightp1 enter the cabin. Itu1 wouldp1 a‹ack John.• However, this strategy leads to overgeneration: any modal (and tense) op-erator introduces a discourse referent, but not all such operators make goodantecedents of namely. Moreover, it is not clear that this discourse referent(p1 in (23)) ranges over denotations of if -clauses, rather than, say, the modalbase.• Summarizing, di‚erent proposals have been made for the source of implicitquestions in discourse in di‚erent constructions:construction source of implicit question:AnderBois andJacobson (2018)namely implicit speci€cational question aboutidentity of d-refAnderBois (2014) sluicing inquisitive contribution of antecedentOnea (2016) namely PPC derived from the antecedent ofnamelyTable 1. Summary of accounts of implicit questionsWhat is the source of the implicit question in the modal case?4 Modal speci€cation and modal restriction4.1 Portner’s semantics of quanti€cational modals• Portner (2009: 218) provides a semantics for quanti€cational modals thatis based on an analysis of adverbs of quanti€cation in the framework ofsituation semantics (Kratzer 2019):(24) a. JsometimesK(α, β) = {s : ∃s′[s′ ≤ s& s′ ∈ counting(α)&∃s′′[s′ ≤ s′′ & s′′ ∈ β]]}b. JcanquantK(α, β) = {s : ∃s′[R(s, ws′) & s′ ∈ counting(α)&∃s′′[s′ ≤ s′′ & s′′ ∈ β]]}Here, counting(α) denotes the set of ‘counting situations ofα’. In Kratzer’s(2019) terms this refers to the situations that exemplify the proposition α,and are maximally self-connected.• In this analysis, quanti€cational modals adopt two properties of adverbs ofquanti€cation that set them apart from other modal operators:1. anti€cational modals quantify over situations, other modal opera-tors quantify over possible worlds.2. anti€cational modals take two arguments instead of just one: themodal prejacent β, and the if -clause α.• Example:(25) a. A spider can be dangerous if it is a‹acked.b. JcanquantK(spider-a‹acked, spider-dangerous)c. informal result of (24): “some counting situation in an accessibleworld of a spider being a‹acked, extends to a situation in whichthe spider is dangerous”• Portner (p. 219) argues that when there is no overt if -clause is present, as in(26) below, the α argument is €lled in by context.(26) Working as a €lmmaker can be taxing.JcanquantK(C,working as a €lmmaker is taxing)5Jos Tellings When if or when specify modals4.2 Modal restriction, speci€cation and exempli€cation: exhaustivity• ‘e standard restrictor analysis of conditionals holds that if -clauses restricta (possibly covert) modal operator (von Fintel 2011). ‘e modal namely con-struction shows that modals and if -clauses can also interact in a di‚erentway.• ‘e meaning of the modal namely construction in (27b) is di‚erent from thatof (27a):(27) a. A spider can be dangerous if it is a‹acked. (=(25))b. A spider can be dangerous, namely if it is a‹acked.One di‚erence is that (27b) carries an implicature of exhaustivity. It im-plicates that the situations in which a spider is dangerous are fully speci€edas the a‹acking situations. (27a) merely says that some situations in whicha spider is a‹acked are ones in which it is dangerous.• ‘e same contrast is observed with temporal quanti€ers such as sometimes,which may make the contrast in (27) clearer.(28) a. I am sometimes sad when it rains.= some raining situations are situations in which I am sadb. I am sometimes sad, namely when it rains.= there are some situations in which I am sad; these situationsare raining situations• A link between namely and exhaustive answers has been made in the nom-inal domain: German na¨mlich provides a complete answer to the implicitquestion that was raised by the antecedent (Onea and Volodina 2011: §4.2;Onea 2016: §6.2.2).3 We have seen something similar in English, for examplein (5b), repeated below, in which an exhaustive list is given:3It is pointed out in these works that the completeness requirement of na¨mlich does not applyto und zwar, another speci€cational adverb in German. It would be interesting to see if/how modalantecedents and if -clause complements are allowed with und zwar, but I haven’t looked into thisyet.(29) IIN may ask for personal data, namely when you: request a catalogor magazine, order books, take our quiz, subscribe to our newslet-ters, or request customer service.In Onea’s work, the completeness requirement is stipulated as a requirementof namely/na¨mlich, but a more explanative account is not given (see Oneaand Volodina 2011: 16 for some discussion).• Compare the reading of namely with for example, which can appear in thesame position:(30) Working as a €lmmaker can be taxing, for example if you’re re-quired to get sleek product shots.‘e complement of for example is conveyed as one of the situations in whichworking as a €lmmaker is taxing, i.e. it is non-exhaustive. See Schwager(2005) for an analysis of the German counterpart zum Beispiel as an operatorthat forces inexhaustive modality.44.3 Licensing namely• I argue that there is a granularity e‚ectwhen it comes to licensing namely:an expression can only serve as the antecedent of namely if it raises an im-plicit question about the identity of an object that is ‘large enough’/clearlyidenti€able.• For example, possible worlds (as in non-quanti€cational modal operators),and events (as in sprouting, recall (16)), are too €ne-grained. However, the‘counting situations’ that quanti€cational modals (24b) and temporal ad-verbs of quanti€cation (24a) quantify over, as well as individuals, are ac-ceptable. See Table 2.• I argue that the question that is raised by the quanti€cational modal in sen-tences such as (27b) asks to specify the α-situations, and the complement ofnamely gives a complete answer, in line with theories that analyze if -clausesas de€nite descriptions (Schlenker 2004).4‘anks to Maria Aloni (p.c.) for referring me to this work.6Jos Tellings When if or when specify modals∃-quanti€cation over: namelypossible? examplespossible worlds N non-quanti€cational modals;(11), (12)events N sprouting; (16)times N simple time adverbials;temporal sproutingindividuals Y regular cases with DPs / d-refscounting situations Y quanti€cational modals /temporal adverbials ofquanti€cationTable 2. Di‚erent levels of granularity(31) Modal speci€cationWorking as a €lmmaker can be taxing, namely if . . . .Q-modalimplicit speci€cationalquestion: What are thecounting situationssuch that β?complete answer• Cases of namely with a non-quanti€cational modal (such as (11) and (12))are typically not judged completely impossible, but degraded. I assume thatin suitable contexts, a non-quanti€cational modal can be (re)interpreted insuch a way that separate circumstances/conditions are made salient. In thatcase, a modal namely construction is licensed.5 Conclusions and some theoretical consequencesOn the distribution of namely• We have seen that namely can take a wider variety of antecedents than justthe DPs that are typically studied. An antecedent such as a modal or anadverbial operator is possible, but only when it is able to raise a questionabout the identity of a suciently large/identi€able formal entity, such as acounting situation.• I hope that in future work, for example using dynamic inquisitive semantics(Dotlacˇil and Roelofsen 2019), the di‚erent notions of implicit questions insluicing, nominal and modal namely-constructions can be given a uni€edtreatment (recall the discussion in section 3.1 and footnote 2 above).On the inquisitive status of modals• ‘ere is some discussion in the literature on inquisitive semantics on the sta-tus of various existential quanti€ers: existential quanti€ers over individualsare inquisitive (Ciardelli et al. 2018), but what about existential quanti€ersover other entities, such as the ones in Table 2 (∃w, ∃e, ∃t, . . . )? (cf. Tellings2019).For example, in AnderBois (2014) it is proposed that existential quanti€ersover events similarly raise an issue about the identity of the event, in orderto account for the behavior of sprouting.• ‘e data analyzed here suggest that the ‘counting situations’ (the informalnatural “units” to count with) are coarse-grained enough to raise an issueabout their identity. ‘ese appear in quanti€cational modals, and in tempo-ral quanti€ers (sometimes, occasionally).• ‘ere is earlier work on the inquisitive status of modals (Ciardelli and Roelof-sen 2018), but this is from a rather di‚erent perspective.7Jos Tellings When if or when specify modalsOn the semantic role of if - and when-clauses• In the data discussed here, if - and when-clauses do not restrict a modal op-erator, but specify it. ‘is corresponds to a di‚erence in meaning (recall(27)/(28)).• Earlier literature on the information-structural properties of conditionalshas pointed out that if -clauses tend to be topics (Haiman 1978; Schi‚rin1992; Ebert et al. 2014), although they can also be focal (Farr 2011, cf. Tellings2018). ‘e construction analyzed here provides yet another example of anif - or when-clause functioning as an answer to a question.A Appendix: some notes on cross-linguistic variation• Many languages have speci€cational adverbs, including Dutch, German,French, Italian, Hungarian, Romanian, Russian, and Chinese (see Onea 2016:§6.1.1 for examples).• Languages di‚er in how many such adverbs they have, as well as in their se-mantic properties (speci€cational/explanative, exhaustive/non-exhaustive,discourse-starting or not, etc.):– Dutch namelijk and German na¨mlich have, in addition to the speci€-cational use, also an explanative use (see footnote 1 above).– Various languages have more than one speci€cational adverb (Germanna¨mlich vs. und zwar ; Hungarian e´spedig vs. megpedig; Dutch namelijkvs. en wel/te weten; French a` savoir vs. c’est-a`-dire). ‘e di‚erent lexicalitems may be associated with di‚erences in distribution and semantics(see e.g. Onea and Volodina 2011; Onea 2016: §6.2 for the German con-trast).• How the modalnamely construction is sensitive to lexical variation of speci-€cational adverbs in languages that have it, is a topic for future investigation.Here I will make a few brief remarks on Dutch and German.• ‘e modal namely construction exists for Dutch and German as well. Beloware examples found online:(32) a. De positie van Major kan nog verder verzwakt worden. Name-lijk als zou blijken dat zijn partij ook bij de Europese verkiezin-gen van gisteren zwaar hee‰ verloren.‘Major’s position can be weakened further. Namely if it turnedout that his party also lost the European elections’.b. Selbst eine wissenscha‰liche Arbeit, deren Inhalte in deinemKopf einfach umwerfend sind, kann noch in die Hose gehen:na¨mlich wenn du sie nicht so formulieren kannst, wie das ander Uni erwartet wird.‘Even a scienti€c paper [. . . ] can go wrong: namely when youcannot formulate it in the way that is expected at university’• Because Dutch and German also have an explanative use of namely, a some-what subtle di‚erence arises between ‘specifying a modal’ and ‘explaining amodal’. ‘e di‚erence, however, is clear due to the di‚erent syntactic prop-erties:(33) a. Een gestolen identiteitskaart kan u veel geld kosten, namelijkwanneer u het slachto‚er wordt van identiteitsdiefstal.‘A stolen ID card may cost you a lot of money, namely whenyou become a victim of identity the‰.’b. Een gestolen identiteitskaart kan u veel geld kosten. Wanneer unamelijk het slachto‚er wordt van identiteitsdiefstal, *(kunnencriminelen geld van uw rekening afschrijven).‘A stolen ID card may cost you a lot of money, because whenyou become a victim of identity the‰, criminals can withdrawmoney from your account’(33a) is an example of speci€cational namelijk, found online. (33b) is theconstructed explanative counterpart. It has namelijk in a di‚erent syntac-tic position, and a main clause is obligatory in the sentence containing thewanneer-clause.8Jos Tellings When if or when specify modalsReferencesAnderBois, S. (2014). ‘e semantics of sluicing: beyond truth conditions. Language,90(4), 887–926. doi:10.1353/lan.2014.0110AnderBois, S., & Jacobson, P. (2018). Answering implicit questions: the case ofnamely. In S. Maspong, B. Stefa´nsdo´‹ir, K. Blake, & F. Davis (eds.), Proceed-ings of SALT 28, pp. 388–408. doi:10.3765/salt.v28i0.4428Brasoveanu, A. (2010). Decomposing modal quanti€cation. Journal of Semantics,27 (4), 437–527. doi:10.1093/jos/‚q008Chung, S., Ladusaw, W. A., & McCloskey, J. (1995). Sluicing and logical form. Nat-ural Language Semantics, 3(3), 239–282. doi:10.1007/BF01248819Ciardelli, I., Groenendijk, J., & Roelofsen, F. (2018). Inquisitive semantics. OxfordUniversity Press.Ciardelli, I., & Roelofsen, F. (2018). An inquisitive perspective on modals and quan-ti€ers. Annual Review of Linguistics, 4, 129–149. doi:10.1146/annurev-linguistics-011817-045626Dotlacˇil, J., & Roelofsen, F. (2019). Dynamic inquisitive semantics: anaphora andquestions. In M. T. Espinal, E. Castroviejo, M. Leone‹i, L. McNally, & C.Real-Puigdollers (eds.), Proceeding of Sinn und Bedeutung 23, pp. 365–382.Universitat Auto`noma de Barcelona, Bellaterra (Cerdanyola del Valle`s).Ebert, C., Ebert, C., & Hinterwimmer, S. (2014). A uni€ed analysis of conditionalsas topics. Linguistics & Philosophy, 37 (5), 353–408. doi:10.1007/s10988-014-9158-4Farr, M.-C. (2011). Focus Inƒuences the Presence of Conditional Perfection: Exper-imental Evidence. In I. Reich, E. Horch, & D. Pauly (eds.), Proceedings of Sinn& Bedeuting 15, pp. 225–239. Saarbru¨cken: Universaar – Saarland UniversityPress.von Fintel, K. (2011). Conditionals. In C. Maienborn, K. von Heusinger, & P. Portner(eds.), Semantics. An International Handbook of Natural Language Meaning.Volume 2, pp. 1515–1538. De Gruyter.Haiman, J. (1978). Conditionals are Topics. Language, 54(3), 564–589.Kratzer, A. (2019). Situations in Natural Language Semantics. In E. N. Zalta (ed.),Œe Stanford Encyclopedia of Philosophy. Summer 2019. Metaphysics ResearchLab, Stanford University. Retrieved from h‹ps://plato.stanford.edu/archives/sum2019/entries/situations-semantics/Onea, E. (2016). Potential questions at the semantics-pragmatics interface. Brill.Onea, E., & Volodina, A. (2011). Between speci€cation and explanation: About aGerman discourse particle. International Review of Pragmatics, 3(1), 3–32.doi:10.1163/187731011X561036Portner, P. (2009). Modality. Oxford University Press.Roelofsen, F., & Van Gool, S. (2010). Disjunctive questions, intonation, and high-lighting. In M. Aloni, H. Bastiaanse, T. de Jager, & K. Schulz (eds.), Logic,language and meaning, pp. 384–394. doi:10.1007/978-3-642-14287-1 39Schi‚rin, D. (1992). Conditionals as topics in discourse. Linguistics, 30, 165–197.Schlenker, P. (2004). Conditionals as de€nite descriptions. Research on Languageand Computation, 2(3), 417–462.Schwager, M. (2005). Exhaustive imperatives. In P. Dekker & M. Franke (eds.), Pro-ceedings of the 15th Amsterdam Colloquium, pp. 233–238. ILLC.Stone, M. (1999). Reference to possible worlds. RuCCS Report 49.Tellings, J. (2018). Œe prosody of additives and the plurality of causation. Colloquiumtalk at Stony Brook University, Nov 2018.Tellings, J. (2019). When-questions and tense in Inquisitive Semantics. In K. Blake,F. Davis, K. Lamp, & J. Rhyne (eds.), Proceedings of SALT 29, pp. 22–42. doi:10.3765/salt.v29i0.45939