Click on the title to expand the page for
more information and link to PDF.
Challenge Problems for a
Theory of Degree Multiplication (with partial answer key)
There are a number of phenomena,
including proportional readings of quantity words and complex unit
expressions like miles per hour
, for which it is tempting
to offer analyses that appeal to fractions involving degrees in the
metalanguage. And yet model-theoretic foundations for quotients of
degrees have not been established within degree semantics. Luckily,
we can look to the study of quantity calculus in the field of
metrology for such foundations. I show how to import a
dimension-centric quantity calculus with multiplication and division
of degrees/quantities into a Montagovian framework, and apply it to
a selection of what I call challenge problems
for a theory
of degree multiplication. Narrowly, the contribution is a
compositional analysis of per
, but the new foundations for
degree semantics have broad potential applicability.
: Coppock, Elizabeth (2021). Challenge Problems for a Theory of Degree Multiplication
(with partial answer key). Talk presented at SALT 31.
Download PDF slides
Granularity in the semantics of comparison
We provide experimental evidence that what matters for definite comparatives
like the bigger circle
in English is not the size of the
comparison class, but the number of degrees it represents. Listeners
flexibly search for a granularity that fits. This paper adds to
the growing evidence that flexible reasoning about granularity has important roles to play in the meaning of gradable expressions.
: Helena Aparicio, Curtis Chen, Roger
Levy and Elizabeth Coppock (2021). Granularity in the semantics of comparison. Talk presented at SALT 31.
Download PDF slides
Mandarin has degree
abstraction after all
Mandarin, along with Japanese, Yorůbá, Mňoré, and Samoan, has been argued to lack so-called `degree abstraction', a configuration at LF involving lambda abstraction over a degree variable. These languages are claimed to have a negative setting for a hypothesized `Degree Abstraction Parameter'. Recent work, however, has argued for degree abstraction in Japanese and Yorůbá, and degree abstraction has been detected in a number of additional languages. Could it in fact be universal? Here, we focus on the case of Mandarin, and argue that Mandarin has degree abstraction too.
We rebut prior arguments for the lack of degree abstraction in Mandarin and discover positive evidence for degree abstraction along the way, considering degree questions, direct measure phrases, subcomparatives and `sub-equatives' (wh-correlatives), scope interactions with modals, negative island effects, attributive comparatives, embedding within clausal comparatives, and superlatives.
Taken together, these results show that degree abstraction is not a parameter along which Mandarin and English vary, and cast further doubt on the existence of the proposed `Degree Abstraction Parameter'.
: Ying Gong and Elizabeth Coppock (2021). Mandarin has degree
abstraction after all. Manuscript, Boston University.
Download PDF manuscript
Differences in implicature across languages stem from differences in salience of alternatives
Scalar implicature depends on the activation of alternatives. For instance, in English, finger
implicates `not thumb', suggesting that thumb
is activated an alternative. Is this because it is more specific (Quantity) and equally short (Manner)? Indeed, toe
doesn't imply `not big toe', perhaps because big toe
is longer. As Larry Horn points out, this Quantity/Manner explanation predicts that if English had the simplex Latin word pollex
meaning `thumb or big toe', then the asymmetry would disappear. But would it suffice for that word to exist in the language, or would the word also have to be sufficiently salient? This thought experiment can be made into an actual one due to the existence of languages like Spanish, which has pulgar
`thumb or big toe' (from pollex
), a non-colloquial form. To gauge the salience of various ways of describing digits, we use a fill-in-the-blank production task with both English, Spanish, Russian, Persian, and Arabic speakers. We then measure the availability of implicatures using a forced choice comprehension task. We find cross-linguistic differences in implicature, and moreover that implicature calculation tracks production probabilities much more closely than mere complexity of the alternatives. A comparison between two Rational Speech Act models -- one in which the speaker perfectly replicates our production data and a standard one in which the speaker chooses based on a standard cost/accuracy trade-off -- shows that comprehension is much more closely tied to production probability than to the mere existence of sufficiently simple alternatives.
: Danielle Dionne and Elizabeth Coppock (2021). Differences in implicature across languages stem from differences in salience of alternatives. Manuscript, Boston University.
Download PDF manuscript
Object agreement in Hungarian: In defense of a semantic solution
This paper contributes another
round in the debate over how to analyze object agreement in
Hungarian. I have previously argued that the choice of conjugation
is determined not by the syntactic category of the object, but
rather on the basis of semantic factors, primarily: on the Lexical
Familiarity Hypothesis (LFH), selected lexical items are assigned a definiteness feature in virtue of a certain type of familiarity presupposition that they carry. Subsequent work has raised challenges for the LFH. This paper considers what would be necessary in order for these challenges can be met. I conclude that the LFH can be defended, if supplemented by a certain set of independently-motivated assumptions. In fact, this theory enjoys certain advantages over the most recent alternative.
: Coppock, Elizabeth (to appear). Object
agreement in Hungarian: In defense of a semantic solution. In
Approaches to Hungarian
17, edited by Tamás Halm,
Balasz Suranyi, and Elizabeth Coppock.
Defining Definiteness in Turoyo
This paper reports on field investigations of the syntactic and semantic factors governing definiteness-marking in Turoyo, an endangered Semitic language. Data collected from translation questionnaires and interviews with native Turoyo speakers shows that Turoyo's definite article has a very wide distribution, covering the full range of uses exhibited by English articles, including giving rise to anti-uniqueness effects with exclusives. They also exhibit double-definiteness uses with demonstratives and possessives, even in non-contrastive environments. The only limit on their distribution is with superlative adjectives, which appear to compete for the article's syntactic position. Based on the broad range of uses, we suggest that Turoyo's definiteness-markers are not 'weak' but 'super-weak' articles that they have familiarity uses because familiarity is a special case of uniqueness.
Yifrach and Elizabeth Coppock (to appear). Defining Definiteness in
Modified numerals: Two routes to ignorance
Modified numerals, such as at least three
or more than five
, are known to sometimes give rise to ignorance inferences, in that their use suggests that the speaker does not have exact knowledge. However, there have been disagreements regarding the nature of these inferences, their robustness and context dependency, and differences between at least
and more than
. We first present a series of experiments which sheds new light on these issues and explains some contradicting reports in the previous literature. Our results show that (a) the ignorance inferences of at least
tend to be more robust than those of more than
, but (b) both modifiers are sensitive to the question under discussion (QUD). We argue that this pattern can be explained if we assume two sources of ignorance: traditional Quantity implicatures, which arise regardless of the modifier but are sensitive to the QUD, and implicatures arising from the maxim of inquisitive sincerity
, which are QUD-insensitive but only appear with at least
Cremers, Alexandre, Elizabeth Coppock, Jakub Dotlacil and Floris
Roelofsen (to appear). Modified numerals: Two routes to
ignorance. Manuscript available on Semantics Archive.
View at Semantics Archive
Nielsen v. Preap / 5th grade grammar v. linguistics / Mass imprisonment v. human rights
In the Supreme Court case Nielsen v. Preap
, ignorance about syntax and semantics led to tragic consequences. The ACLU lawyer defending thousands of non-citizens from being rounded up and put into prison indefinitely by ICE let it come across that her argument rested on the false premise that adverbs can modify nouns. The textualists claimed victory, even though the humane reading of the text was the literal one in this case. The final decision rested crucially on this error on her part, and was buffered by a misunderstanding about how definite descriptions work. The dissent failed to articulate a convincing rebuttal, making spurious ref- erence to passive voice. This case clearly shows how staggeringly consequential linguistic knowledge can be.
Coppock, Elizabeth (2021). Nielsen v. Preap / 5th grade grammar
v. linguistics / Mass imprisonment v. human rights. In
Proceedings of the Linguistic Society of America
797--805. DOI: 10.3765/plsa.v6i1.5015
Access PDF on LSA website
| Watch video on Youtube
Universals in Superlative Semantics
This paper reports on the results of a broad cross-linguistic study on the semantics of quantity words such as many in the superlative (e.g. most
). While some languages use such a form to express both a relative reading (as in Gloria has visited the most continents
) and a proportional reading (as in Gloria has visited most continents
), the vast majority do not allow the latter, though all allow the former. It is argued that a degree-quantifier analysis of quantity superlatives is best suited to explain why proportional readings typically do not arise for them. Based on morphosyntactic evidence, two alternative diachronic pathways through which proportional quantifiers may develop from quantity superlatives are identified.
Coppock, Elizabeth, Elizabeth Bogal-Allbritten and Golsa Nouri-Hosseini (2020). Universals in Superlative
Quantification, degrees, and beyond in Navajo
This paper explores the interaction of degree and quantification in Navajo. We argue that Navajo comparative standard markers should be analyzed as quantifiers over degrees despite the language apparently only allowing phrasal (non-clausal) standards of comparison. Our primary evidence comes from novel exploration of superlative meaning in Navajo, which is conveyed via the combination of a comparative standard marker and an existential affix that obligatorily takes lowest scope, a configuration which our analysis of standard markers makes possible. We develop a compositional analysis that resolves the quantificational meanings of standard markers with non-clausal standards. By positing degree quantifiers in Navajo, we challenge earlier analyses which took Navajo to lack expressions of quantificational determiner type. We posit that such meanings exist only in the domain of degrees because degree arguments are the only type of argument position in Navajo that are not obligatorily saturated by verbal prefixes.
: Bogal-Allbritten, Elizabeth and
Elizabeth Coppock (2020). Quantification, degrees and beyond
in Navajo. In Interactions of Degree and Quantification
ed. Peter Hallman, pp. 121--162. Leiden: Brill.
Informativity in Image Captions vs. Referring Expressions
At the intersection between computer vision and natural language processing, there has been recent progress on two natural language generation tasks: Dense Image Captioning and Referring Expression Generation for objects in complex scenes. The former aims to provide a caption for a specified object in a complex scene for the benefit of an interlocutor who may not be able to see it. The latter aims to produce a referring expression that will serve to identify a given object in a scene that the interlocutor can see. The two tasks are designed for different assumptions about the common ground between the interlocutors, and serve very different purposes, although they both associate a linguistic description with an object in a complex scene. Despite these fundamental differences, the distinction between these two tasks is sometimes overlooked. Here, we undertake a side-by-side comparison between image captioning and reference game human datasets and show that they differ systematically with respect to informativity. We hope that an understanding of the systematic differences among these human datasets will ultimately allow them to be leveraged more effectively in the associated engineering tasks.
: Coppock, Elizabeth, Danielle
Liu, and Derry
Wijaya (2020). Informativity in Image Captions vs. Referring
Expressions. In Proceedings of the Probability and Meaning
Conference (PaM 2020)
, Gothenburg, Sweden,
ed. Christine Howes, Stergios Chatzikyriakidis, Adam Ek, and Vidya Somashekarappa, pp. 104--108. Association for Computational Linguistics.
Access PDF on ACL website
Most vs. the most in languages
where the more means most
This paper focuses on languages in which a superlative interpretation
is typically indicated merely by a combination of a definiteness
marker with a comparative marker, including French, Spanish,
Italian, Romanian, and Greek ('DEF+CMP languages'). Despite ostensibly using definiteness markers to form the superlative, superlatives are not always definite-marked in these languages, and the distribution of definiteness-marking varies across languages. Constituency structure appears to vary across languages as well. To account for these patterns of variation, we identify conflicting pressures that all of the languages in consideration may be subject to, and suggest that different languages prioritize differently in the resolution of these conflicts. What these languages have in common, we suggest, is a mechanism of Definite Null Instantiation for the degree-type standard argument of the comparative. Among the parameters along which languages are proposed to differ is the relative importance of marking uniqueness vs. avoiding determiners with predicates of entities that are not individuals.
Coppock, Elizabeth and Linnea Strand (to appear). Most
vs. the most
where the more
. In Ana Aguilar-Guevera,
Julia Pozas Loyo and Violeta Vázquez Rojas Maldonado (eds.),
Definiteness Across Languages
, 271--417. Berlin: Language Science
Download PDF manuscript
It's not what you expected! The surprising nature
of cleft alternatives in French and English
While much prior literature on the meaning of clefts -- such as the
English form 'It is X that Z' -- concentrates on the nature and
status of the exhaustivity inference ('nobody/nothing other than X Z'), we report on experiments examining the role of the doxastic status of alternatives
6 on the naturalness of c'est-clefts in French and and it-clefts in English. Specifically, we study the hypothesis that clefts indicate a conflict with a doxastic commitment held by some discourse participant. Results from naturalness tasks suggest that clefts are improved by a property we term 'contrariness' (along the lines of Zimmermann, 2008). This property has a gradient effect on felicity judgments: the more strongly interlocutors appear committed to an apparently false notion, the better it is to repudiate them with a cleft.
Destruel, Emilie, David I. Beaver and Elizabeth Coppock (2019). It's not what you expected! The surprising nature
of cleft alternatives in French and English.
Frontiers in Psychology
10. DOI: 10.3389/fpsyg.2019.01400.
Published online at Frontiers in Psychology
Quantity superlatives in Germanic, or, Life on the fault line between adjective and determiner
This paper concerns the superlative forms of the words many
, and little
, and their
equivalents in other Germanic languages (German, Dutch, Swedish,
Norwegian, Danish, Dalecarlian, Icelandic, and Faroese). It
demonstrates that every possible relationship between definiteness
and interpretation is attested. It also demonstrates that agreement
mismatches are found with relative readings and with proportional
readings, but different kinds of agreement mismatches in each
case. One consistent pattern is that a quantity superlative with
adverbial morphology and neuter singular agreement features is used
with relative superlatives. On the other hand, quantity superlatives
with proportional readings always agree in number. I conclude that
quantity superlatives are not structurally analogous to quality
superlatives on either relative or proportional readings, but they
depart from a plain attributive structure in different ways. On
relative readings they can be akin to pseudopartitives (as in a
cup of tea
), while proportional readings are more closely
related to partitives (as in a piece of the cake
specifically, I suggest that the agreement features of a superlative
exhibits depend on the domain from which the target is drawn (the
target-domain hypothesis). When the target is a degree, as it is
with adverbial superlatives and certain relative superlatives,
default neuter singular emerges. Definiteness there is driven by the
same process that drives definiteness with adverbial
superlatives. With proportional readings, the target argument of the
superlative is a subpart or subset of the domain indicated by the
substance noun, hence number agreement. Subtle aspects of how the
comparison class and the superlative marker are construed determine
definiteness for proportional readings.
Coppock, Elizabeth (2019). Quantity superlatives in Germanic, or,
Life on the fault line between adjective and determiner. To appear in
Journal of Germanic Linguistics
Journal of Germanic Linguistics
Forces at the interface of gradability and quantification
This paper ties together four cross-linguistic generalizations: (i) proportional readings for quantity superlatives are typologically marked; (ii) adverbial superlatives have only relative readings; (iii) quantity superlatives agree in number with the noun on proportional, but not relative readings; (iv) adverbial morphosyntax can be used with quantity superlatives on relative readings. We propose that the commonalities between quantity and adverbial superlatives are due to the fact that comparison is being made among degrees or events rather than individuals, and offer a compositional account.
Coppock, Elizabeth and Elizabeth Bogal-Allbritten (2018). Forces at the interface of gradability and quantification. In
Proceedings of SALT 28
, edited by Sireemas Maspong,
Brynhildur Stefánsdóttir, Katherine Blake, and Forrest Davis, pp. 747--767.
Download PDF manuscript
This paper presents and advocates an approach to the semantics of opinion statements, including matters of personal taste and moral claims. In this framework, 'outlook-based semantics', the circumstances of evaluation are not composed of a possible world and a judge (as in 'world-judge relativism'); rather, outlooks
replace possible worlds in the role of circumstance of evaluation. Outlooks are refinements of worlds that settle not only matters of fact but also matters of opinion. Several virtues of the framework and advantages over existing implementations of world-judge relativism are demonstrated in this paper. First, world-judge relativism does not actually explain the 'disagreement' of 'faultless disagreement', while a straightforward explanation suggests itself in outlook-based semantics. Second, outlook-based semantics provides an account of subjective attitude verbs that can capture lack of opinionatedness. Third, outlook-based semantics unproblematically explains the connection-building role of aesthetic discourse and the group-relevance of discretionary assertions, while capturing the same effects in world-judge relativism obviates the purpose of the judge parameter. Finally, because the proposed circumstances of evaluation (outlooks) are entirely analogous to possible worlds, the framework is easy to use and extend.
Coppock, Elizabeth (2018). Outlook-based semantics. Linguistics
and Philosophy 41: 125--164.
View on Linguistics and Philosophy website
| Download PDF
The proper treatment of egophoricity in Kathmandu Newari
We develop a theory of so-called 'conjunct-disjunct marking', also
known as 'egophoricity', in Kathmandu Newari. The signature pattern
of egophoricity looks a bit like person agreement: In declaratives,
there is a special marker that goes on first person verbs, but not
second or third person (e.g. 'I drank-EGO too much'). But in
interrogatives, the same marker goes on second person (e.g. 'Did
you-EGO drink too much?'). This is called interrogative
. Egophoric marking also interacts interestingly with the
presence of evidential markers, and comes with an implication of
knowing self-reference (emphasized in Newari by a restriction to
volitional action). Our paper discusses two previous approaches,
which we label indexical and evidential, and motivate our account,
which we label egophoric. Along the way, we develop a theory of how
de se attitudes are communicated.
Coppock, Elizabeth and Stephen Wechsler (2018). The proper
treatment of egophoricity in Kathmandu Newari. In Minyao Huang and
Kasia M. Jaszczolt (eds.), Expressing the Self: Cultural Diversity and
. Oxford: Oxford University Press, pp. 40-57.
Absolut superlativ i samtida sprĺkbruk
In this article we report on a corpus study of elative superlatives
in contemporary Swedish. Elative superlatives differ from ordinary
superlatives in that no direct comparison with other referents is
involved. Instead a referent is said to have the property expressed
by the adjective to a very high degree. In Swedish, elative
superlatives are formally distinct from ordinary superlatives (they
lack the post-nominal clitic article) which means that they can be
found in corpus searches. We show that elative superlatives have
expressive function and are typically used in emphatic assertions
which are intended to make the strongest possible claim in a given
situation. Elative superlatives are used in all grammatical
functions but with slightly different implicational
properties. Contrary to what has been assumed, elative superlatives
are not limited to fixed expressions and formal written
language. Creative uses abound in blog texts and sports
Engdahl, Elizabet and Elizabeth Coppock (2017). Absolut
superlativ i samtida sprĺkbruk. Sprĺk och stil
Download PDF manuscript (in Swedish)
Storyboards vs. Picture-aided translation: A case study on the typology of comparison
We present a comparison of a storyboard method with a method we call
, where text is presented along
with pictures. Results show that picture-aided translation has
certain advantages. We also discuss best practices for picture
design, based on our process of developing and testing a story with
accompanying images for the purpose of studying quantity comparison.
Nouri-Hosseini, Golsa, and Elizabeth Coppock (2017). Storyboards vs. Picture-aided translation: A case study on the typology of comparison. Workshop on Elicitation Tools for Linguistic Description and Typology, University of Paris Diderot, November 2017.
Download PDF slides
Clefts: Quite the contrary!
We give experimental evidence for the hypothesis that clefts are more
felicitous when used to correct an incorrect belief on the part of
Destruel, Emilie, David Beaver and Elizabeth Coppock (2017). Clefts: Quite the contrary! In
Robert Truswell (ed.), Proceedings of Sinn und Bedeutung
. Edinburgh: University of Edinburgh.
Implicatures of modified numerals: Quality or quantity?
We propose a new analysis of modified numerals that allows us to: (i)
predict ignorance with respect to the prejacent of at least
(and thereby avoid to Bernard Schwarz's recent criticism of Coppock and Brochhagen
2013), (ii) get a three-way contrast between superlative modifiers,
comparative modifiers, and numerals, without appeal to a two-sided
analysis of numerals, and (iii) avoid the prediction that at
should produce quantity implciatures when only
is not a grammatical alternative. With it, we reconcile
Westera and Brasoveanu's (2014) findings with the achievements of
the Coppock and Brochhagen account, bring that work in line with
recent theorizing in inquisitive semantics using downward-closed
possibilities, and show that inquisitive sincerity can interact with
Horn-based quantity in a non-trivial way, something that may be
fruitful to consider in other domains as well.
Ciardelli, Ivano, Elizabeth Coppock, and Floris Roelofsen (2017). Implicatures of modified numerals: Quality or quantity? In
Robert Truswell (ed.), Proceedings of Sinn und Bedeutung
. Edinburgh: University of Edinburgh.
Proportional implies relative: A
We give evidence from a geographically, genetically, and typologically
diverse set of languages (drawn from 26 different language families
and every continent) for the following typological universal:
Regardless of the morphosyntactic strategy used by a language to
form superlatives, if superlative morphosyntax can be applied to
, then the result can be used to
express a relative reading (as in Hillary has visited the most
continents (out of everyone)
) but not necessarily a
proportional reading (as in Hillary has visited most of the
). Thus, no language deploys the regular superlative
for the proportional but not the
relative reading. We also give a rough estimate of how rare
proportional readings for quantity superlatives are: about
10%. Nevertheless, we show that proportional readings arise with a
diverse set of strategies for forming superlatives, and discuss the
cases of Basque, Hausa, and Georgean (the last of which is most
Coppock, Elizabeth, Golsa Nouri-Hosseini, Elizabeth Bogal-Allbritten
and Saskia Stiefeling (to appear). Proportional implies relative: A
typological universal. Proceedings of the Linguistic Society of
Link to paper at LSA Proceedings website
Superlative modifiers as modified superlatives
The superlative modifiers at least
and at most
quite famous, but their cousins at best
, at the
, at the highest
, etc., are less
well-known. This paper is devoted to the entire family. New data is
presented illustrating the productivity of the pattern, identifying
a generalization delimiting it, and showing that the cousins, too,
have the pragmatic effects that have attracted so much attention to
at least and at most. To capture the productivity, I present a new
decomposition of at least into recombinable parts. Most notable is
-component (silent in some languages), which takes
advantage of the comparison class argument of the superlative to
produce the set of possibilities involved in the ignorance
implicatures that superlative modifiers are known for. A side-effect
is a new view on gradable predicates, accounting for uses like
88 degrees is too hot
Coppock, Elizabeth (2016). Superlative modifiers as modified
superlatives. Proceedings of SALT
26, 471-488. doi:http://dx.doi.org/10.3765/salt.v26i0.3822
View PDF online
Why possessives should not be discussed at this conference
Schoorlemmer (1998) makes the following generalization: adjectival
possessives (in e.g. Italian) can be either definite or indefinite,
but determiner-like possessives (in e.g. English) are always
definite. Many others have argued for definiteness of
determiner-like possessive constructions. Partee & Borschev (2003)
say the prenominal genitive in English seems to combine the basic
genitive [the post-nominal form] with an implicit definite article,
while Vikner & Jensen (2002) argue that a possessive behaves as if
it had an implicit definite article]. On the other hand, Haspelmath
(1999) and Peters & Westerstĺhl (2013) argue against inherent
definiteness of determiner-like possessives. We will support this
latter position, which points towards a uniform analysis of
determiner-like and adjectival possessives. Specifically, we suggest
that the cross-linguistic data motivating Schoorlemmer's
generalization can be explained without positing inherent
definiteness of possessives, using the framework introduced by
Coppock and Beaver (2015).
David Beaver and Elizabeth Coppock (2016). Why possessives should not
be discussed at this conference. Talk presented at the Definiteness
Across Languages workshop in Mexico City, June 2016.
Download PDF slides
Sophisms and insolubles
Just prior to the spread of universities across Europe in the
fourteenth century, a systematic method for training the
minds of young future leaders to think rationally began to
crystallize through the practice of logical disputations. In
fourteenth-century Oxford, before earning a Bachelor of
Arts, a student was required to earn the title of
. As such, he was allowed to
participate in structured disputations involving a
respondent and an opponent, and would have learned the art
of considering a sentence called a sophism
) against a hypothetical scenario or given
set of assumptions, called a casus
Latin. Typically, it was not trivial to decide whether the
sentence was true or false, and arguments could be made on
both sides. Sophisms thus presented a puzzle to be solved.
This paper gives an overview of the topics dealt with in
the sophism literature, including the
interpretation of so-called 'syncategorematic' terms such as
, semantic paradoxes including
the liar paradox ('insolubles'), mathematical physics, and
questions related to knowledge and belief. The paper also
discusses comparisons between this medieval enterprise and
modern formal semantics vis-a-vis the treatment of presupposition,
the distinction between object language and
representation language, and proof-theoretic semantics.
Yrjönsuuri, Mikko and Elizabeth Coppock (2016). Sophisms and
Insolubles. In Catarina Dutilh Novaes and Stephen Read (eds.),
The Cambridge Companion to Medieval Logic
265-289. Cambridge: Cambridge University Press.
Download PDF manuscript
Quasi-Definites in Swedish: Elative
Superlatives and Emphatic Assertion
This paper analyzes nominal phrases in Swedish with a definite article but no definite suffix on the head noun, which we call quasi-definites (e.g. det största intresse
'the greatest interest'). These diverge from the usual 'double definiteness' pattern where the article and the suffix co-occur (e.g. det största intresse-t
'the greatest interest-def'). We give several diagnostics showing that this pattern arises only with superlatives on an elative ('to a very high degree') interpretation, and that quasi-definites behave semantically as indefinites, although they have limited scope options and are resistant to polarity reversals. Rather than treating the article and the suffix as marking different aspects of definiteness, we propose that both are markers of uniqueness and that the definite article functions within the adjectival phrase and combines with a predicate of degrees rather than individuals in this construction. The reason that quasi-definites do not behave precisely as ordinary indefinites has to do with their pragmatics: Like emphatic negative polarity items, elative superlatives require that the assertion be stronger (approximately more surprising) than alternatives formed by replacing the highest degree with lower degrees, and have a preference for entailment scales.
Coppock, Elizabeth and Elisabet Engdahl (2016). Quasi-Definites in
Swedish: Elative Superlatives and Emphatic Assertion. Natural
Language and Linguistic Theory
Download PDF manuscript
Novelty and familiarity for free
We argue that a uniqueness-based theory of definiteness gives you
novelty/familiarity for free, as long as you have some way of
tracking discourse referents, and some way of restricting
descriptions to the property of identity with a referent. The rest
Coppock, Elizabeth and David Beaver (2015). Novelty and familiarity
for free. Proceedings of the 20th Amsterdam Colloquium
University of Amsterdam.
Definiteness and Determinacy
This paper distinguishes between
definiteness and determinacy. Definiteness is seen as a
morphological category which, in English, marks a (weak) uniqueness
presupposition, while determinacy consists in denoting an
individual. Definite descriptions are argued to be fundamentally
predicative, presupposing uniqueness but not existence, and to
acquire existential import through general type-shifting operations
that apply not only to definites, but also indefinites and
possessives. Through these shifts, argumental definite descriptions
may become either determinate (and thus denote an individual) or
indeterminate (functioning as an existential quantifier). The latter
option is observed in examples like "Anna didn't give the only
invited talk at the conference", which, on its indeterminate
reading, implies that there is nothing in the extension of "only
invited talk at the conference". The paper also offers a resolution
of the issue of whether possessives are inherently indefinite or
definite, suggesting that, like indefinites, they do not mark
definiteness lexically, but like definites, they typically yield
determinate readings due to a general preference for the shifting
operation that produces them.
Coppock, Elizabeth and David Beaver (2015). Definiteness and
Determinacy. Linguistics and
View PDF online (open access!)
Completely Bare Swedish Superlatives
This paper shows that Swedish differs from both German and English with respect to the distribution and interpretation of definiteness-marking on superlatives: Bare degree and amount superlatives unambiguously receive a relative interpretation, definite-marked amount superlatives are unambiguously 'proportional' (although they do not always carry a 'more than half' interpretation), and definite-marked degree superlatives can have an absolute or a relative reading. We show that an analysis based on movement of the superlative morpheme accounts well for the Swedish pattern but does not provide the tools for a cross-linguistically valid framework, failing in particular to account well for relative readings in conjunction with definiteness-marking. We therefore propose an alternative, non-movement approach building on a very recent treatment of the superlative morpheme, giving it access to a contrast set and an association relation. The crucial difference between Swedish on the one hand and English and German on the other hand is proposed to lie in whether the association relation is saturated through semantic composition or by context.
Coppock, Elizabeth and Christian Josefson (2015). Completely Bare Swedish
Superlatives. In Eva Csipak and Hedde Zeijlstra (eds.), Proceedings
of Sinn und Bedeutung 19
Minimal Sufficiency Readings in Conditionals
We discuss minimal sufficiency readings of exclusives like
, as in Just the thought of him sends shivers down
, which does not mean the same thing as Only the
thought of him sends shivers down my spine
. We provide a set of
diagnostics for identifying minimal sufficiency readings in
conditionals and in simple clauses, and identify a generalization as
to where the latter type appear: only in arguments that have a
'causer' thematic role. For this reason, we see minimal sufficiency
readings in conditionals as basic, and provide an analysis of them
building on Kratzer's notion of a modal base.
Coppock, Elizabeth and Anna Lindahl (2015). Minimal Sufficiency
Readings in Conditions. In Christopher Brown, Qianping Gu, Cornelia
Loos, Jason Mielens, and Grace Neveu (eds.), Proceedings of the 15th Texas Linguistic Society
Truth Value Judgments vs. Validity Judgments
This paper undertakes a direct comparison between two methodologies for getting at semantic intuitions: (i) validity judgments, where subjects judge the validity of arguments, e.g. There are three bananas; therefore there are at least three bananas
, and (ii) picture verification tasks (also known as `truth judgment tasks'), in which one sees a picture of three bananas and judges a statement like There are at least three bananas
. It has been suggested that validity judgment tasks are more sensitive to ignorance implicatures than picture verification tasks, but these two methods have not been compared directly using comparable stimuli. The present work aims to close that gap. The results show that validity judgment tasks do not in fact robustly pick up on ignorance implicatures, so they cannot be relied upon for that, although both validity judgment tasks and truth value judgment tasks are sensitive to violations of particularly strong pragmatic requirements. In general, the two kinds of tasks gave quite similar results. This raises the question why validity judgment tasks sometimes pick up on ignorance implicatures and sometimes do not.
Coppock, Elizabeth (2015). Truth Value Judgments vs. Validity
Judgments. In Christopher Brown, Qianping Gu, Cornelia
Loos, Jason Mielens, and Grace Neveu (eds.), Proceedings of the
15th Texas Linguistic Society
A Superlative Argument for a Minimal Theory of Definiteness
This paper argues that the
distinction between absolute and relative readings of superlatives
supports a distinction between definiteness and determinacy.
Previous work has suggested that definite superlative noun phrases
like "the fewest letters" in "Gloria received the fewest letters"
(relative superlative DPs) are semantically indefinite. This paper
argues that such DPs are definite (presupposing uniqueness) but not
determinate (denoting an individual). We provide new evidence that
they are not determinate, and undertake a critical review of the
evidence that has previously been used to argue for their
indefiniteness, arguing that it is consistent with the hypothesis
that they are definite. We argue furthermore that a movement
analysis of relative superlatives is not consistent with a treatment
of the determiner as definite. We therefore offer an analysis of
relative superlative DPs on which they are definite but
indeterminate, and the superlative morpheme is interpreted in situ.
Coppock, Elizabeth and David Beaver (2014). A Superlative Argument
for a Minimal Theory of Definiteness. In Todd Snider (ed.), Proceedings of SALT 24
Paul Elbourne, Definite Descriptions
Elbourne's book sets a new
standard for the analysis of definite descriptions, meeting an
impressive set of subtle targets. But the theory of partiality and
presupposition needs some work, and certain ideas were too hastily
dismissed, including dynamic semantics and the notion that definite
descriptions are basically predicative.
Coppock, Elizabeth (2014). Paul Elbourne, Definite
. Nordic Journal of Linguistics
View at Nordic Journal of Linguistics
Principles of the Exclusive Muddle
The words only
, adjectival only
, and alone
are members of a unified class -- the class of exclusives -- in a
sense that this paper makes precise. The most famous representative of
this class is only
, for which at least 27 distinct lexical
entries have been given. This paper situates only
context of its lexical relatives in English, accounting for a number
of equivalences and non-equivalences between sentences involving
and ones involving other exclusives. We propose that
what unifies the words mentioned above is that they concern an upper
bound on the viable answers to the current question under discussion,
and signal that a lower bound on them is taken for granted. These two
criteria are encapsulated in a lexical entry schema for exclusives,
which accommodates two main points of variation: semantic type (within
the class of modifiers), and constraints on the current question under
discussion. We propose 17 different specific instantiations of the
schema for the exclusives listed above.
Coppock, Elizabeth and David Beaver (2014). Principles of the
Exclusive Muddle. Journal
Journal of Semantics
Download PDF manuscript
Raising and Resolving Issues with Scalar Modifiers
This paper argues that the
superlative modifiers at least
and at most
lower and upper bounds, respectively, on the true answers to the
question under discussion (QUD), and that they are
with an analysis of only
on which it presupposes a lower
bound on the QUD and asserts an upper bound on the QUD, our analysis
of superlative modifiers yields a unified picture of these scalar
items, and accounts for the connection between them. This analysis
also successfully accounts for their truth conditions,
focus-sensitivity, distribution, and interaction with modals.
Analyzing at least
and at most
as inquisitive in the
inquisitive semantics sense yields a satisfactory account of the fact
that, in contrast to corresponding sentences without such items, they
do not give rise to quantity implicatures, and yet they do give rise
to ignorance implicatures, in contrast to comparatives. Superlative
modifiers thus both depend on the QUD for their interpretation, and
raise issues for discussion.
Coppock, Elizabeth and Thomas Brochhagen. Raising and Resolving
Issues with Scalar Modifiers. Semantics and Pragmatics
6(3): 1--57. 2013.
online at S&P
Diagnosing Truth, Interactive Sincerity, and Depictive Sincerity
This paper presents experimental
evidence from picture-verification tasks that the superlative
modifiers at least
and at most
give rise to an
, rather than signalling ignorance as an
entailment. More surprisingly, this paper reports on heretofore
unnoticed behavior of at most
: while There are at most 4
is consistently judged as true in a scene with 4
butterflies, There are at most 5 butterflies
is not. The
corresponding contrast is not found for at least
interpret this using the notion of "highlighting" from inquisitive
semantics, and draw the broader conclusion that some pragmatic effects
are so strong that they can affect truth/falsity judgments.
Coppock, Elizabeth and Thomas Brochhagen (2013). Diagnosing Truth,
Interactive Sincerity, and Depictive Sincerity. In Todd Snider
(ed.), Proceedings of SALT 23
View PDF online
A Semantic Solution to the Problem of Object Agreement in Hungarian
Hungarian verbs are sensitive to the
definiteness of their object, but the definiteness of the object is
not a completely reliable indicator of the subjective/objective
alternation. The current state of the art is an analysis due
originally to Bartos (2001), according to which the syntactic status
of the object is the determining factor: If it is a DP or larger
phrase, then the objective conjugation is used; otherwise the
subjective conjugation is used. This purely syntactic analysis is
problematic for several reasons pointed out by Coppock & Wechsler
(2012). This paper presents a semantic solution that overcomes these
difficulties: If the referential argument of a phrase is lexically
specified as familiar, then the phrase bears the feature [+DEF], and
in this case it triggers the objective conjugation. The analysis is
implemented in a fragment of Hungarian that includes a wide range of
nominal expressions, using a compositional version of van der Sandtian
Discourse Representation Theory.
Coppock, Elizabeth (2013). A Semantic Solution to the Problem of Object Agreement in Hungarian. Natural Language Semantics
View PDF online
Download PDF manuscript
Agreement Between Scylla and Charybdis
This is a short introduction to
agreement in HPSG, with emphasis on semantic issues, meant to serve as
an introduction to Steve Wechsler's chapter on Swedish "pancake"
sentences. It reviews the fact that this theory of agreement avoids
the Scylla of a purely syntactic theory, while at the same time
avoiding the Charybdis of a purely semantic one. The ontological
status of HPSG's indices
, which play a critical role in
achieving this delicate balance, is discussed, and readers are
encouraged to consider adopting this theory of agreement whether or
not they are working in HPSG per se.
Coppock, Elizabeth (2013). Agreement Between Scylla and Charybdis. In The Core and the Periphery: Data-Driven Perspective
on Syntax Inspired by Ivan A. Sag
, ed. Philip Hofmeister and
Elisabeth Norcliffe, pp. 65-70. Stanford: CSLI Publications.
Turkic Non-Inflectional Plurals and Feature Bundling
We defend the bundling
for agreement, which entails that agreement on a
target that contains person will always reflect the person feature of
the controller. A putative counterexample to this hypothesis comes
from Sakha exceptional case marking constructions, where verbs appear
to agree in number with a raised accusative object while ignoring
person. We argue that this case is not a genuine counterexample,
however, because the agreement marker in question is not a person
marker. Evidence for this comes from its distribution (optionality,
co-occurrence with local persons), separate morphological expresssion
from person, and history.
Coppock, Elizabeth, Hyun-Jong Hahm and Stephen Wechsler
(2013). Turkic Non-Inflectional Plurals and Feature Bundling. In Umut Özge (ed.), Proceedings of
the 6th Workshop on Altaic Formal Linguistics
, MIT Working
Papers in Linguistics volume 67, pp. 53-64. 2013.
This paper offers a unified analysis
of the exclusives mere
. We analyze both
in terms of questions under discussion
modeled as structures defined over sets of alternatives, where the
alternatives are non-standard in two respects. First, as required by
the analysis of mere, there are free variables in the alternatives.
Second, the alternatives can be ranked by relations other than
entailment; in this sense, both mere
scalar both in their positive component ("at least X") and in their
negative component ("at most X"). With evidence from negation, reason
clauses, and emotive factive predicates, we argue furthermore that the
negative component of mere
contributes to the at-issue
meaning, while the positive component is presupposed, as has been
previously argued for only
. However, mere
differ in scope, as evidenced by differences in
interpretation and NPI licensing. Based on the NPI licensing
properties of mere
, we argue that it has two uses, one that
attaches to properties and one that attaches to generalized
quantifiers. The two uses are unified under one abstract lexical entry
schema that can be extended to only
as well. This schema can
be seen as a first step towards establishing, in general terms, the
core meaning for exclusives.
Coppock Elizabeth and David Beaver (2013). Mere
-ology. In Alternatives in Semantics
, edited by Anamaria Falaus, pp. 150--173. New York: Palgrave. 2013.
Weak Uniqueness: The only difference between definites and indefinites
We argue that predicative
is an identity function that is defined for predicates
that satisfy weak uniqueness
: if there is an F
there is only one. Predicative definites do not presuppose existence,
as evidenced by anti-uniqueness effects, for example the fact that
`Scott is not the only author of Waverley
' implies that there
is more than one author of Waverley
. The definite and
indefinite articles are bothargued to be identity functions on
predicates, differing only inthat the latter lacks the weak uniqueness
presupposition. Furthermore, the meaning of argumental definites and
indefinites can be derived from the predicative meanings using the
same general mechanisms that introduce existence. Existence is
generally at-issue with argumental indefinites and presupposed with
argumental definites. However, we observe that anti-uniqueness effects
arise with argumental definites as well, which we take to show that
existence can be at-issue with argumental definites.
In Proceedings of SALT 22, ed. Anca Chereches, Neil Ashton and David Lutz, pp. 527--544. eLanguage.net. 2012.
It-clefts are IT (inquiry terminating) constructions
We analyze the semantics of a range
of constructions which we refer to as Inquiry Terminating (IT)
constructions. In English, these include it
exclusives such as only, just
Despite their differences, IT constructions have much in common. We
claim they are always focus-sensitive, have closely related semantics,
and have a uniform discourse function: they always mark utterances
that give a complete answer to what the speaker takes to be the
Current Question (CQ). We give a new account of the meaning of clefts
that captures both their similarities and their differences to other
In Proceedings of SALT 22, ed. Anca Chereches, Neil Ashton and David Lutz, pp. 441-460. eLanguage.net. 2012.
Exclusivity, Uniqueness, and Definiteness
This paper argues for two main
claims: (i) The definite article is initially predicative and
contributes a weak uniqueness
presupposition, which is
logically independent of existence. Only in argument position does a
definite (or indefinite) article signal existence. (ii) A distinction
is to be drawn between pure exclusive adjectives (adjectival
) and cardinality adjectives (single
can function as both, and can also be
used as a quantifier.With these assumptions, we can explain
anti-uniqueness effects that only
to in predicative definite descriptions, and the fact that
but not only
is compatible with the indefinite
article. The distinction between exclusive and singular-cardinality
adjectives has broader empirical consequences as well; exclusive
adjectives are compatible with plurals but singular-cardinality
adjectives are not, and cardinality adjectives can modify superlatives
but exclusive adjectives cannot.
In Empirical Issues in Syntax and
Semantics 9, pp. 59-66, ed. Christopher Pińon. 2012.
This paper develops a type of dynamic
semantics in which contexts include not only information, but also
questions, whose answers are ranked by strength. The questions can be
local to the restrictor of a quantifier, and the quantifier can bind
into them. The proposed framework satisfies several desiderata arising
from quantificational expressions involving exclusives (e.g.
only, just, mere,
allowing: (i) presupposed questions; (ii) presuppositional constraints
on the strength ranking over the answers to the question under
discussion; (iii) quantificational binding into such presupposed
questions; and (iv) compositional derivation of logical forms for
In Logic, Language and Meaning: 18th Amsterdam
Colloquium, edited by Maria Aloni, Floris Roelofsen, Galit
Weidman Sassoon, Katrin Schulz and Matthijs Westera, pp. 291-300,
Berlin: Springer. 2012.
The objective conjugation in Hungarian: Agreement without phi features
Verbal agreement is normally in
person, number and gender, but Hungarian verbs agree with their
objects in definiteness instead: a Hungarian verb appears in the
when it governs a definite object. The
sensitivity of the objective conjugation suffixes to the definiteness
of the object has been attributed to the supposition that they
function as incorporated object pronouns (Szamosi 1974, den Dikken
2006), but we argue instead that they are agreement markers
registering the object's formal, not semantic, definiteness. Evidence
comes from anaphoric binding, null anaphora (pro
extraction islands, and the insensitivity of the objective conjugation
to any of the factors known to condition the use of affixal and clitic
pronominals. We propose that the objective conjugation is triggered by
a formal definiteness feature and offer a grammar that determines, for
a given complement of a verb, whether it triggers the objective
conjugation on the verb. Although the objective conjugation suffixes
are not pronominal, they are thought to derive historically from
incorporated pronouns (Hajdu 1972), and we suggest that while
referentiality and phi-features were largely lost, an association with
topicality led to a formal condition of object definiteness. The
result is an agreement marker that lacks phi-features.
Natural Language and Linguistic Theory
30(3), pp. 699-740.
Focus as a case position in Hungarian
I argue that in Hungarian,
accusative case can be assigned by a verb to a noun phrase, triggering
definiteness agreement, purely in virtue of the fact that the noun
phrase serves as the focus in the same clause as the verb. As shown by
data from incorporation, long-distance constructions, binding, and
depictives, accusative case-marked focus-raised subjects are not
objects of the verb -- not thematic objects or even athematic objects
-- so their only link to the verb that assigns case to them is via
information structure. Based on the Case-Assignment Generalization --
A verb agrees in definiteness with a noun phrase in Hungarian if and
only if it assigns accusative case to it -- I have argued that the
same holds for accusative case-marked focus-raised objects that agree
in definiteness with their are also examples where accusative case is
assigned to a nominal purely in virtue of its information- structural
status. Thus, both subject and object focus raising exemplify
case-assignment to a noun phrase whose sole link to the case-assigning
verb is through information structure. This means that the focus
position in Hungarian is one to which accusative case can be assigned.
In Discourse and Grammar: A Festschrift in Honor of
Valéria Molnár, edited by Johan Brandtler, David Hĺkansson,
Stefan Huber, and Eva Klingvall, pp. 161-178, Lund: Centre for
Languages and Literature.
We propose a unified analysis of
exclusives, taking into account NP- and VP-modifying only
well as just
and the adjectival exclusives mere
, and exclusive
Using paraphrases with at most
and at least,
that exclusives uniformly signify a presupposed lower bound and an
ordinary content upper bound on the true alternative answers to the
current question under discussion, thus extending Beaver and Clark 2008.
We propose that exclusives vary along two parameters: (i) the
ontological type of their arguments; (ii) constraints on the question
under discussion. Due to variation in the type parameter, exclusives
exhibit different scopes, leading to different NPI licensing properties.
To formalize our analysis, we introduce a dynamic semantics that treats
questions under discussion as part of the context and allows for binding
into these questions.
| In Proceedings of SALT 21, edited by Neil Ashton, Anca Chereches, and David Lutz, pp. 197-217, eLanguage.net. 2011.
Less-travelled paths from pronoun to agreement
Building on Bresnan and Mchombo's
(1987) theory that the transition from pronoun to agreement marker
constitutes the loss of a PRED 'pro' specification on an affix, we
explore the idea that the historical path from pronoun to agreement
marker can involve the loss of person and number feature
specifications as well. We apply this idea to object agreement in the
Uralic languages, with particular attention to Ostyak and Hungarian,
and propose that person and number specifications on object agreement
affixes, historically derived from bound pronouns, were lost
independently at different stages. We then consider the more general
hypothesis that the special distribution of person agreement can be
explained as a consequence of its historical origin in incorporated
pronouns, with loss of the person feature as a complicating factor.
Preliminary typological evidence supports this view over Baker's
(2008) theory of person agreement.
In Proceedings of LFG '10, Miriam Butt and Tracy King, pp. 165-185, Stanford: CSLI Publications. 2010.
Variation in the Iraq vowel: Conservatives vs. liberals
To determine whether phonological
variables are a potential resource for the expression of political
identity, this article examines the second vowel of Iraq
. In addition
to being part of a politically significant place-name, Iraq
particularly well-suited to index political identity due in part to
the ideological association between the "foreign (a)" variable with
correctness and educatedness in U.S. English (Boberg 1997).
Specifically, Iraq's second vowel appears to index political
conservatism when produced as /ae/ and political liberalism when
produced as /a:/. Results from an analysis of the U.S. House of
representatives show that Republicans are significantly more likely
than Democrats to use /ae/, even controlling for regional accent.
American Speech 85: 91-102. 2010.
A translation from logic to English with dynamic semantics
This paper presents a procedure for translating standard predicate logic into English. The procedure generates both referring expressions and non-referring expressions, including both referential and bound variable anaphora. Non-referring expressions correspond to short-term discourse referents, which present a special set of challenges for a natural language generation system: (i) they have limited ÔlifespansŐ and (ii) the determiner with which they are introduced (every, some, any, no) is sensitive to the logical context. Our system addresses these challenges using dynamically updated information states.
In Daisuke Bekki, Yohei Murakami and Eric McCready (eds.), 6th International Workshop on Logic and Engineering of Natural Language Semantics (LENLS 6), 197-216. Berlin: Springer. 2010.
Parallel encoding of alternatives in sentence production: Evidence from syntactic blends
Using a large, newly available corpus of spontaneously uttered
syntactic blends (e.g. "cast into question" from the targets "call into
question" and "cast into doubt") and a new method of speech error
analysis, two hypotheses regarding grammatical encoding are compared:
the single-buffer hypothesis, according to which alternative
formulations of the message are encoded in the same memory buffer,
potentially sharing representations, and the multiple-buffer
hypothesis, according to which alternative formulations are
independently grammatically encoded in separate buffers. Randomly
generated, unattested blends were found to be reliably distinguishable
from blends attested in the corpus, based on the degree to which they
adhere to syntactic alignment constraints, controlling for other
important factors. This main finding suggests that elements in similar
syntactic positions across plans compete for the same slot, supporting
the single-buffer hypothesis.
Language and Cognitive Processes 25(1): 38-49. 2009.
The Logical and Empirical Foundations of Baker's Paradox
My dissertation addresses questions of syntactic productivity of the
following form: Based on independent (e.g. semantic) properties of a
given word, can it be predicted whether the word may occupy a given
syntactic position? For example, can it be predicted from the meaning
of disappear that it cannot function transitively (*He disappeared
the rabbit)? Why is mere restricted to prenominal position
(*The child is mere)? The putative existence of arbitrary
exceptions has been used as an argument in favor of a usage-based
approach to grammar and acquisition. This work shows that the
argument from arbitrary exceptions does not hold water, as the
phenomena in question turn out to be explainable in terms of deeper, more
general principles upon closer inspection.
In Search of the True Periphery: Why Culicover's "Odd Prepositions" Aren't that Odd
This paper argues, contra Culicover (1999), that individual prepositions do not differ arbitrarily in their ability to strand (Who are you talking to?) or pied-pipe (To whom are you talking?). This removes the argument that the features [STRAND] and [PIEDPIPE] must be set for each word individually on the basis of positive experience (nor is there even any evidence that such features exist). Prepositions do not differ arbitrarily in whether they precede or follow their argument, either. [PRECEDE NP] and [FOLLOW NP] are likewise not features that must be set individually on the basis of positive experience, if they exist. This leads to the suggestion that the learner is not, as Culicover argues, Ňconservative and attentiveÓ in assigning syntactic properties to words.
In Proceedings of the 33rd Meeting of the Berkeley Linguistic Society. Berkeley: Berkeley Linguistics Society. 2007.
ISIS: It's not a disfluency, but how do we know that?
This paper argues that the ISIS construction (e.g. "The thing is, is that...") is not a disfluency, based primarily on acoustic comparison between this construction and other repetitions of "is" in the Switchboard corpus.
In Proceedings of the 32nd Meeting of the Berkeley Linguistic Society. Berkeley: Berkeley Linguistics Society. 2006.
Gapping: In Defense of Deletion
I propose a deletion analysis of
Gapping (e.g. John likes caviar, and Mary beans
), and argue
that such an analysis makes better predictions than one on which
Gapping is Across-The-Board movement.
In In Mary Andronis, Christopher Ball, Heidi Elston,
and Sylvain Neuvel (eds.), CLS 37: The Main Session. Papers from
the 37th Meeting of the Chicago Linguistic Society, pp. 133-148.
Chicago: Chicago Linguistic Society. 2001.