A Note on Quasi-Presuppositions and Focus M´arta Abrus´an University of Oxford January 31, 2011
1
Introduction
It has been long observed that certain modifiers trigger a presupposition-like inference (cf. Bellert 1977, Linebarger 1980, Simons 2001, Schlenker 2008, Chemla 2009a,b). The primary examples of such examples are VP-adverbs and other predicate modifiers. An example is the adverb late below: (1)
a. b. c.
John did not come late implies: John came If John came late, the party was a disaster implies: John came Did John come late? implies: John came
All the utterances above imply that the speaker believes that John came, which suggests that the inference that arises might be a presupposition. Nevertheless, one would be hard pressed to claim that the adverb late itself lexically presupposes the truth of the proposition that it modifies, nor is it likely that the expression came late itself presupposes came. How does then the inference arise? Three possibilities have been suggested in the literature (cf. Simons 2001, Schlenker 2008, Chemla 2009a,b). One possibility is that the inference arises as an (embedded) scalar implicature. A second possibility, argued for by Simons (2001) and Schlenker (2008), is that these inferences are after all a type of presupposition, termed ‘quasi-presupposition’ by Schlenker (2008). Under this view, presuppositions or presupposition-like phenomena can arise compositionally, via a pragmatic triggering mechanism, hence no lexical encoding of the presupposition is necessary. ‘Quasi-presuppositions’ are then inferences that exhibit Many thanks to Mats Rooth, two anonymous reviewers, Emmanuel Chemla, Philippe Schlenker, Benjamin Spector and the participants of the Aspects of Prosody Workshop in Oxford for helpful comments. The research reported here was supported by the ESF (Euryi grant to P. Schlenker) and The Mellon Foundation. All remaining errors are my own.
1
a presupposition-like projection behavior, but are weaker than standard presuppositions. A third possibility is that the inference arises because the modifiers in question are focused, and the inference we observe is the givenness effect of focus. Experimental results of Chemla (2009a,b) show that inferences that arise from adverbial modification are in between presuppositions and implicatures with respect to how strongly they project from the scope of negation and some downward entailing quantifiers, and with respect to response times. Compare the results in the scope of the quantifier no. In this context, presuppositions of classical triggers such as know project universally, as shown in (2a). The scalar implicature of all in (2b) is only predicted to lead to an existential inference, which was experimentally confirmed too. Adverbial modifiers show an inference that is between the previous two: universal inferences are accepted at a smaller rate than for the factive presupposition of know, but at a higher rate than for scalar implicatures. (2)
a.
b. c.
None of these 10 students knows that his father is going to receive a congratulation letter. implies: Each of these students’ fathers is going to receive a congratulation letter. None of these 10 students read all the books. implies: (At least) one of them read some of the books. None of these ten boys came late weakly implies: All these ten boys came
These results suggest that the inference that arises from adverbial modifiers is not simply an implicature. But they are still compatible with both of the second two approaches: the idea that the inference arises as a type of presupposition and that it arises via focus.1 This squib presents an argument in support of the view that the inference that arises with some cases of adverbial modification is the givenness effect of focus. The main source of evidence comes from Hungarian, a language which marks focus syntactically. Thus it is possible to tell without prosodic or contextual information whether or not a constituent is focused. When adverbs or other modifiers are not focused in Hungarian, no ‘quasipresupposition’ effect is generated. I also adduce further evidence from English suggesting that focus is the culprit in this language as well. Finally, I note that the idea that the quasipresupposition effect arises from focus is consistent with the observation that the inference created by modifiers is weaker than that created by most standard presupposition triggers. In the remainder of this squib I use the expression quasi-presupposition as a nontheoretical, descriptive term for the inference that arises with certain cases of adverbial modification. 1
The comparison with the projection properties of presuppositions might be more complicated if it turned out that not all presupposition triggers project as strongly in quantificational contexts as know, cf. Charlow (2009) for arguments in this direction. This does not change the point about implicatures.
2
2
Focus and Quasi-Presuppositions: Hungarian
Focus is usually understood to be the part of an utterance that conveys the new or highlighted information. In English, focus is standardly assumed to be marked by heavy stress or pitch accent on the focused constituent (cf. Chomsky 1971, Rochemont 1986 and many others). In the following question-answer pair the focus of the answer is on the subject DP. Main stress falls on the subject DP, Peter. (3)
a. b.
Who ate the cake? [Peter]F ate the cake.
In Hungarian focus is marked by a special word order: the Hungarian focus construction (cf.(4)). The focused constituent appears in the left periphery of the clause, immediately ´ Kiss 1994, 1998, 2002, Szendr˝oi preceding the finite V (cf. Szabolcsi 1981, Szabolcsi 1994, E. 2001). (4)
[P´etert]F ismerte meg Mari. Peter.acc know prt Mari ‘Mari got to know [Peter]F ’
A diagnostic for Hungarian focus is that the aspectual particle of the verb (meg in the above example) has to appear after the verb. Thus (4) contrasts minimally with (5) in which the preverbal constituent cannot be understood as focused. (5)
P´etert megismerte Mari. Peter.acc prt.know Mari ‘Mari got to know Peter.’
So, while in English focus is marked solely by prosodic means, by main stress, in Hungarian focus is always marked syntactically, by a special word order.2 Given this, if the verb has an aspectual particle, one can tell with certainty whether or not an adverbial modifier is focused in Hungarian simply by looking at the syntactic form of the sentence. The adverb hangosan ‘loudly’ is focused in (6b) and not focused in (6a), which is marked by the position of the aspectual particle el : (6)
a. b.
P´eter hangosan el´enekelte a Himnuszt Peter loudly prt.sang the anthem.acc P´eter [hangosan]F ´enekelte el a Himnuszt Peter loudly sang prt the anthem.acc ‘Peter sang the anthem [loudly]F ’
2
Note that this does not mean that Hungarian focus is not also marked by prosody: In particular, Szendr˝ oi 2001 argued that the rigid syntactic requirement on Hungarian focus follows from the specifics of Hungarian prosody, in which main stress always has to fall on the pre-verbal position.
3
Importantly, in Hungarian quasi-presupposition facts only arise when the adverb is focused, as in (7b). When it is not focused, as in (7a), no such effect can be observed. (7)
a.
b.
K´etlem, hogy P´eter hangosan el´enekelte a Himnuszt doubt.1sg that Peter loudly prt.sang the anthem.acc does not imply: Peter sang the anthem K´etlem, hogy P´eter [hangosan]F ´enekelte el a Himnuszt doubt.1sg that Peter loudly sang prt the anthem.acc ‘I doubt that Peter sang the anthem [loudly]F ’ implies: Peter sang the anthem
Further, (as Simons 2001 also notes for English) the effect disappears if something else in the sentence is focused. In example (8) the subject Peter is focused, and the quasipresupposition effect does not arise: (8)
K´etlem, hogy [P´eter]F ´enekelte el hangosan a Himnuszt doubt.1sg that Peter sang prt loudly the anthem.acc ‘I doubt that [Peter]F sang the anthem loudly’ does not imply: Peter sang the anthem implies: someone sang the anthem loudly
The same effect can be observed not only with manner adverbs, but also with quantificational adverbs such as k´etszer ‘twice’ and high adverbs such as sz´and´ekosan ‘deliberately’: (9)
a.
b.
c.
K´etlem, hogy P´eter k´etszer/sz´and´ekosan elt¨orte a l´ab´at doubt.1sg that Peter twice/deliberately prt.broke the leg.3sg.acc does not imply: Peter broke his leg K´etlem, hogy P´eter [k´etszer/sz´and´ekosan]F t¨orte el a l´ab´at doubt.1sg that Peter twice/deliberately broke prt the leg.3sg.acc ‘I doubt that Peter broke his leg [twice/deliberately]F ’ implies: Peter broke his leg K´etlem, hogy [P´eter]F t¨orte el k´etszer/sz´and´ekosan a l´ab´at doubt.1sg that Peter broke prt twice/deliberately the leg.3sg.acc ‘I doubt that [Peter]F broke his leg twice/deliberately’ does not imply: Peter broke his leg implies: Somebody broke his leg twice/deliberately
Focusing does not interact with the lexical interpretation of the adverbs hangosan ‘loudly’, k´etszer ‘twice’ and sz´and´ekosan ‘deliberately’ in the examples above. But interestingly, in Hungarian many low (manner) adverbs have a strong preference for being focused, while ´ Kiss (2009). many high adverbs seem to have a preference against being focused, cf. E. This might account for the fact noted by Bellert (1977) that high adverbs tend not to show the presupposition effect. Further, some adverbs, e.g. gyorsan ‘quickly’ have both a high and low reading (cf. Eszes 2009). When this adverb is not focused, as in (10a), it has a strong preference for a high reading under which it was the the decision making of Peter 4
that was quick, but not necessarily the climbing itself. When the adverb is focused, the low (manner) reading is more salient. As expected, it is only the low, focused reading in (10b) that gives rise to a presupposition. (10)
a.
b.
K´etlem, hogy P´eter gyorsan felm´aszott a f´ara doubt.1sg that Peter quickly prt.climbed the tree.on does not imply: Peter climbed the tree K´etlem, hogy P´eter [gyorsan]F m´aszott fel a f´ara doubt.1sg that Peter quickly climbed prt the tree.on ‘I doubt that Peter climbed the tree [quickly]F ’ implies: Peter climbed the tree
The Hungarian facts thus clearly show that the quasi-presupposition effect we are observing is the effect of focus. But there is suggestive evidence arguing that the quasipresupposition effect is due to focus in English as well. First, the examples that illustrate the phenomenon always have the adverb in a sentence final position, as in (1). Given that in English sentences with neutral intonation the final phonological phrase bears the nuclear accent and it can therefore easily be understood as focused (cf. e.g. Herburger 2000), it is likely that in the relevant English examples what researchers have been identifying as the quasi-presupposition effect was also the effect of focus. Second, as noted by Simons (2001), the effect disappears in English as well if another constituent is focused in the sentence. Instead, what seems to project is the givenness effect generated by the focused constituent, in the example below that Bill did something fast: (11)
Bill didn’t [run]F fast.
Another argument from English, which I owe to an anonymous reviewer, is that if there are two adverbs in the sentence, what seems to project is everything that precedes the adverb in the final position. This can be explained easily if we assume that the last modifier, being in sentence final position, associates with focus and thus requires everything else to be given. (12)
Bill didn’t run to the park quickly.
In Hungarian, normally only one constituent can occupy the focus position. Therefore if there are two modifiers, only one of them can be focused. Similarly to English, what we observe is that what projects like a presupposition is the givenness effect associated with the focused constituent: (13)
a.
b.
K´etlem, hogy P´eter [hangosan]F ´enekelte el a Himnuszt a parkban doubt.1sg that Peter loudly sang prt the anthem.acc the park.in ‘I doubt that Peter sang the anthem in the park [loudly]F ’ implies: Peter sang the anthem in the park K´etlem, hogy P´eter [a parkban]F ´enekelte el a Himnuszt hangosan doubt.1sg that Peter the park.in sang prt the anthem.acc loudly 5
‘I doubt that Peter sang the anthem loudly [in the park]F ’ implies: Peter sang the anthem loudly somewhere It is mysterious how a view that claims that certain modifiers give rise to a presuppositionlike effect can account for the facts that involve a focus-driven shift of what is presupposed. Its proponents would have to argue that the quasi-presupposition effect of modification can be overridden by focus, and explain why and how that is predicted. However, on the account that reduces the quasi-presupposition effects to focus, these shifts are just what is predicted. If the quasi-presupposition effect is to do with focus, how does this square with Chemla’s (2009b) results? It is well known that the background inference of focus seems to project out of presupposition holes (cf. Jackendoff 1972, Kratzer 1989, Kadmon 2001, Geurts and van der Sandt 2004): (14)
John didn’t marry [Sue]F implies: John married someone
Projection of the givenness requirement of focus from the scope of quantifiers has not been much discussed in the literature. In Hungarian focus seems to show a universal projection pattern under the quantifier no 3 : (15)
a.
b.
Egyik fi´ u se [Bert´anak]F k´erte meg a kez´et One boy neither Berta.dat asked prt the hand.3sg.acc ‘None of the boys proposed to [Berta]F ’ implies: They all proposed to someone Egyik fi´ u se k´erte meg a kez´et Bert´anak One boy neither asked prt the hand.3sg.acc Berta.dat ‘None of the boys proposed to Berta’ implies: nothing
The preference for a universal projection in the case of (15a) is very much in line with Chemla’s (2009b) findings. The explanation for the slightly weaker acceptance rate might follow from the fact that focus places a weaker requirement on the context, as discussed below, or from the fact that Chemla tested written French examples in which focus was not marked.
3
Antecedent Facts
Proponents of the analysis according to which quasi-presuppositions are pragmatically triggered by some hitherto undiscovered special mechanism have pointed out that the inference that results from modifiers in English is weaker than standard presuppositions 3
Negation also requires an inverse verb-particle order in Hungarian, which is why the particle is still postverbal in (15b), despite there not being any focus.
6
(cf. Simons 2001 and Schlenker 2008)4 . Experimental results reported in Chemla (2009a,b) seem to support this claim. In this section I observe that this state of affairs is compatible with the focus analysis as well. In the preceding discussion I have referred to the pragmatic effect that focus places on context as the ‘givenness effect of focus’. Traditionally focus was assumed to carry a genuine existential presupposition (cf. Chomsky 1971, Williams 1980, etc.), but in more recent mainstream semantic approaches to focus only a weaker condition on context was assumed (Rooth 1992, Schwarzschild 1999, e.g.), whereby focus only requires the non-focussed material to be given in the preceding (possibly embedded) context. This position has been again challenged by Geurts and van der Sandt (2004) and Abusch (2010), mainly based on various projection facts associated with focus. In this paper I wish to remain agnostic about whether focus triggers an existential presupposition or only a givenness requirement, as both positions would support the conclusion of the previous section. Nevertheless, it seems fair to say that focus places a weaker requirement than many of the traditional presupposition triggers on the context (cf. Rooth 1996, Geurts and Sandt 2004, Schwarzschild 2004, B¨ uring 2004), whether it can be called presupposition or not. The interesting point to note then is that quasi-presuppositions place a similar requirement on the context as focus, namely they show a relative freedom of associating with antecedents. Rooth (1996, 1999) gives the following examples to differentiate the effect of focusing on discourse and existential presupposition. He starts from the rather uncontroversial assumption that it-clefts give rise to existential presuppositions. Cf. the contrast among the following: (16)
a. b. c.
A: Did anyone win the football pool this week? B: Probably not, because it’s unlikely that [Mary]F won it, and she is the only person who ever wins B’: #Probably not, because it’s unlikely that it’s Mary who won it, and she is the only person who ever wins
Suppose that B knew that Mary made a silly bet. The existential inference of focus and the existential presupposition of the it-cleft both conflict with the first part of B’s utterance. But only the example with the it-cleft leads to a contradictory assertion. This argues that focus either does not presuppose the truth of the existential inference, or at least it it has a more volatile presupposition than it-clefts.5 4
Hungarian focus is usually assumed to give rise to somewhat stronger effects than English focus, ´ Kiss namely it tends to be interpreted exhaustively, rather like clefts in English, cf. Szabolcsi (1981), E. ´ (2010). See also Kenesei (1986) and E. Kiss (2008) for a claim that Hungarian focus triggers an existential presupposition. However, the argument from focus goes through even if Hungarian focus (and quasi presuppositions) triggers stronger effects than the ones described in this section: what is important is that when the adverb is not focused, there residue does not project either. Nevertheless, it is worth mentioning that some researchers have recently challenged the idea that Hungarian focus has a strong semantics (cf. Wedgewood 2005, Onea and Beaver 2009, Szendr˝oi and Br´ody 2010) and argued that the effects observed are pragmatic. 5 For some speakers, Hungarian shows the same pattern wrt. this example. However, according to an
7
In examples similar to Rooth’s, the quasi-presupposition examples pattern together with the focus examples above: (17)
Context: B saw all the people who came in the finish during the first hour or so–i.e. the fast people a. A: Did John run? b. B: Probably not, because obviously he did not run fast, and he is always among the fast ones.
It seems that the quasi-presupposition that arises in the example above is not globally accommodated, unlike the presupposition of the it-cleft in Rooth’s example. A different example to show that focus does not trigger an existential presupposition was given in Schwarzschild (2004): (18)
If the Health Ministry has discovered a cure for cancer, they are probably still testing it, and if [Merck]F has discovered a cure for cancer, they are probably working on a marketing plan.
The above example does not commit us to the view that someone has discovered a cure for cancer. Instead, the givenness requirement of focus can be satisfied by the linguistic material in the antecedent of the first conditional. This is unlike the presupposition of the definite description the cure they discovered in (19) below, the presupposition of which cannot be satisfied by the antecedent of the first conditional and so its presupposition is accommodated globally. (19)
If the Health Ministryi has discovered a cheap cure for cancer, they are probably still testing it, and if the cure theyi discovered is expensive, they are probably working on a funding plan.
Contrast this now with the example below: (20)
If the Health Ministry made a breakthrough discovery of a cure for the deadly flu, then everyone survived, but if they discovered a cure for the flu late, everyone died.
anonymous reviewer, many native speakers do not find (ib) acceptable: (i)
(=(16)) a. A: Megnyerte valaki a tot´ ot a h´eten? ‘Did anyone win the toto this week?’ b. B: K´etlem, mert nem val´ osz´ın˝ u hogy [Mari]F nyerte meg, ´es ˝o az egyetlen aki nyerni szokott. ‘I doubt it, because it is not likely that [Mari]F won it, and she is the only one who wins’. c. B’: #K´etlem, mert nem val´ osz´ın˝ u Mari volt az aki megnyerte, ´es ˝o az egyetlen aki nyerni szokott. ‘I doubt it, because it is not likely that it was Mari who won it, and she is the only one who wins’.
8
Analogously to Schwarzschild’s first example, the above does not commit us to the view that the health ministry discovered any cure. So it seems that the quasi-presupposition of the second clause can be satisfied by the antecedent of the first conditional, thus its projection properties are like that of focus, and unlike that of definite descriptions. The above data seem to show that indeed the inference triggered by modifiers seems to be weaker than the presupposition triggered by definite descriptions or clefts. They are not in themselves arguments for the focus account, as not all presupposition triggers require the strong contextual licensing conditions that definite descriptions or clefts impose (cf. e.g. Stalnaker 1973; Beaver 2004; Simons 2001; Abbott 2006; Abusch 2010).6 But the examples do show that the focus analysis is just as consistent with the observation that the inference is a weak one, as a conversationally based triggering story would be.
4
Conclusion
This squib gave arguments for the position that so called quasi-presupposition effects of adverbial modification arise from focus. The main argument came from Hungarian, a language that marks focus syntactically, and it was shown that such effects only arise if the adverb or modifier is focused. Evidence from English corroborating this position was brought up as well. Finally, it was observed that the focus analysis is compatible with the observation according to which quasi presuppositions are weaker than standard presuppositions: the contextual requirements of quasi-presuppositions are similar to that of focus, which have been argued to be themselves weaker than standard triggers. These facts suggest that an explanation based on focus is highly plausible.
References Abbott, B. (2006). Where have some of the presuppositions gone? In B. J. Birner and G. Ward (Eds.), Drawing the boundaries of meaning: Neo-Gricean studies in pragmatics and semantics in honor of Laurence R. Horn, pp. 1–20. Philadelphia: John Benjamins. Abusch, D. (2010). Presupposition triggering from alternatives. Journal of Semantics 27 (1), 37–80. Beaver, D. (2004). Have you noticed that your belly button lint colour is related to the colour of your clothing? In R. Bauerle, U. Reyle, and T. E. Zimmerman (Eds.), Presupposition: Papers in Honor of Hans Kamp. 6
The following examples were offered by an anonymous reviewer, showing that the factive implication of discover can be canceled in similar environments as the ones cited above: (i)
A: Was Mrs. B a British spy? B: I don’t think so, because the CIA didn’t discover she WAS, despite extensive efforts to prove it.
(ii)
Cancer is probably incurable. But if the Health Ministry has discovered that it can be cured with high doses of illegal drugs, they probably wouldn’t tell anyone.
9
Bellert, I. (1977). On semantic and distributional properties of sentential adverbs. Linguistic Inquiry 8 (2), 337–351. B¨ uring, D. (2004). Focus suppositions. Theoretical Linguistics 30 (1), 65–76. Charlow, S. (2009). Strong “Predicative Presuppositions”. Ms., NYU. Chemla, E. (2009a). An experimental approach to adverbial modification. In U. Sauerland and K. Yatsushiro (Eds.), Semantics and Pragmatics: From Experiment to Theory, pp. 249–263. Palgrave Macmillan New York. Chemla, E. (2009b). Presuppositions of quantified sentences: experimental data. Natural Language Semantics 17 (4), 299–340. Chomsky, N. (1971). Deep structure, surface structure, and semantic interpretation. Semantics: An interdisciplinary reader in philosophy, linguistics and psychology, 183–216. ´ Kiss, K. (1994). Sentence structure and word order. In F. Ferenc Kiefer and K. E.Kiss ´ E. (Eds.), The syntactic structure of Hungarian, pp. 1–90. San Diego, California: Academic Press Inc. ´ Kiss, K. (1998). Identification focus and information focus. Language 74, 245–273. E. ´ Kiss, K. (2002). The Syntax of Hungarian. Cambridge: Cambridge University Press. E. ´ Kiss, K. (2008). Deriving the properties of structural focus. In The 18th International Congress E. of Linguistics, 2008, Korea University, Seul. The Linguistic Society of Korea. ´ Kiss, K. (2009). Scalar adverbs in and out of focus. In K. E. ´ Kiss (Ed.), Adverbs and Adverbial E. Adjuncts at the Interfaces. Berlin: Mouton de Gruyter. ´ Kiss, K. (2010). Structural focus and exhaustivity. In C. F´ery (Ed.), Information Structure. E. Theoretical, Typological and Experimental Perspectives. Oxford: Oxford University Press. ´ Kiss (Ed.), Eszes, B. (2009). Aspect and adverb interpretation–the case of quickly. In K. E. Adverbs and Adverbial Adjuncts at the Interfaces . Berlin: Mouton de Gruyter. Geurts, B. and R. Sandt (2004). Interpreting focus again. Theoretical Linguistics 30 (1), 149–161. Geurts, B. and R. van der Sandt (2004). Interpreting focus. Theoretical Linguistics 30 (1), 1–44. Herburger, E. (2000). What counts: focus and quantification. The MIT Press: Cambridge, MA. Jackendoff, R. S. (1972). Semantic Interpretation in Generative Grammar. Cambridge, Mass: The MIT Press. Kadmon, N. (2001). Formal Pragmatics: Semantics, Pragmatics, Presupposition, and Focus. Blackwell Publishers. Kenesei, I. (1986). On the logic of word order in hungarian. In W. Abraham and S. de Mey (Eds.), Topic, Focus and Configurationality, pp. 143–159. John Benjamins: Amsterdam. Kratzer, A. (1989). An investigation of the lumps of thought. Linguistics and Philosophy 12 (5), 607–653. Linebarger, M. (1980). The grammar of negative polarity. Ph. D. thesis, Massachusetts Institute of Technology. Onea, E. and D. Beaver (2009). Hungarian Focus is not Exhausted. In Proceedings of SALT, Volume 19. Rochemont, M. (1986). Focus in generative grammar. John Benjamins, Amsterdam. Rooth, M. (1992). A theory of focus interpretation. Natural Language Semantics 1 (1), 75–116. Rooth, M. (1996). Focus. In S. Lappin (Ed.), The Handbook of Contemporary Semantic Theory, pp. 271–297. Oxford: Blackwell. Rooth, M. (1999). Association with focus or association with presupposition, pp. 232–244. Cambridge University Press. Schlenker, P. (2008). Be Articulate: A Pragmatic Theory of Presupposition Projection. Theoret-
10
ical Linguistics 34 (3), 157–212. Schwarzschild, R. (1999). Givenness, AvoidF and other Constraints on the Placement of Accent. Natural Language Semantics 7 (2), 141–177. Schwarzschild, R. (2004). Focus interpretations: Comments on Geurts and van der Sandt (2004). Theoretical Linguistics 30 (1), 137–147. Simons, M. (2001). On the conversational basis of some presuppositions. In J. B. Hastings, R. and S. Zvolenszky (Eds.), Semantics and Linguistic Theory, 11. Stalnaker, R. C. (1973). Presuppositions. Journal of Philosophical Logic 2 (4), 447–457. Szabolcsi, A. (1981). The semantics of topic-focus articulation. In T. Janssen and M. Stokhof (Eds.), Formal methods in the study of language, pp. 513–41. Amsterdam: Matematisch Centrum. Szabolcsi, A. (1994). All quantifiers are not equal. Acta Linguistica Hungarica 42, 171–187. Szendr˝oi, K. (2001). Focus and the Syntax-Phonology Interface. Ph. D. thesis, University College London. Szendr˝oi, K. and M. Br´ ody (2010). Exhaustive focus is an answer. Ms., UCL and HAS. Wedgewood, D. (2005). Shifting the Focus. From Static Structures to the Dynamics of Interpretation. Elsevier. Williams, E. (1980). Remarks on Stress and Anaphora. Journal of Linguistic Research 1 (3), 1–16.
11
