Workshop on States
Organizers: Alexis Wellwood, Daniel Altshuler, Ashley Atkins
Speakers: Bridget Copley, Ashley Atkins, Daniel Altshuler, Daniel Skibra, Roger Schwarzschild, Fabrizio Cariani, Alexis Wellwood, Rebekah Baglini, Itamar Francez, Andrew Koontz-Garboden
Northwestern University, May 5-6, 2017
Speakers: Bridget Copley, Ashley Atkins, Daniel Altshuler, Daniel Skibra, Roger Schwarzschild, Fabrizio Cariani, Alexis Wellwood, Rebekah Baglini, Itamar Francez, Andrew Koontz-Garboden
Northwestern University, May 5-6, 2017
Join us for an interdisciplinary workshop focusing on states.
Bridget Copley: How to Do Things When You Don't Have Any Energy
The common-sense notion of a state entails a lack of energy. Since we also have an intuition that energy is needed for causation, we should expect states not to be able to cause things to happen. Yet there is ample linguistic evidence suggesting that states do cause things to happen. In this talk I enumerate some of the ways states can be causal despite their lack of energy, and analyze a particular class of causal states in English have-causatives and futurates. The syntax-semantics interface guides us to an understanding of such states, which turn out to be at once transcendant and familiar. Throughout, the question of what is visible and what is invisible to the grammar is addressed.
Ashley Atkins: Back to the Futurate
Futurates have been described as sentences that convey that a future eventuality is planned, scheduled, or otherwise determined, but where this future orientation isn't obviously due to any overt linguistic element. Consider the futurate sentences, 'Mary is crossing the Atlantic tomorrow' and 'Mary crosses the Atlantic tomorrow,' for example. We appreciate that they both convey that a cross-Atlantic passage is due to take place tomorrow. But why aren't these sentences heard as contradictory? Why isn't the first heard as saying, for example, that a cross-Atlantic passage that occurs tomorrow is ongoing at the present time? 'Mary is crossing the Atlantic' is certainly heard as saying that a cross-Atlantic passage is presently ongoing. (Consider now the implausibility of assuming that 'tomorrow' contributes the meaning that such an event is due to take place the day after the day of utterance.) The central challenge posed by futurates is taken to be to account for the emergence of their future-oriented interpretations in the absence of any overt future-oriented meanings.
In her foundational work on futurate meaning, Bridget Copley pursued three assumptions that guided her toward a response to this challenge. The first is that these future-oriented interpretations are due to a covert modal meaning. The second is that this sort of meaning, though not uniform across various futurate expressions (e.g., progressive and simple) is linked to a system of meaning, and in particular, to the imperfective system, widely assumed to be a major natural language modal system. And third, any appeal to differences in event structure—the assumption that futurates are events with planning stages, for example—obscures rather than illuminates the nature of futurate meaning. I want to pursue a different course in my presentation, taking these three guiding assumptions as points of departure. I want to take seriously the possibility that the future orientation of futurates isn't due to any covert meaning (and so not to a covert modal meaning), that one of the important consequences of this assumption is that the imperfective system is not a modal system, and that appealing to structural differences—in particular, differences in the interactions between stative and telic meanings—can yield a uniform account of futurate meaning.
Daniel Altshuler: Distinguishing State and Time Anahora: A Look at Now
Since the work of Kaplan (1989) and Kamp (1971), the philosophical and linguistic tradition has treated now as a pure indexical: its linguistic meaning determines the referent on an occasion of use. In particular, its referent is the time of utterance on every occasion of use. There have been several challenges for this view, starting from the observation that now often occurs in narratives where it seems to pick out a time in the past (Dry 1979, Kamp & Rohrer 1983, Kamp & Reyle 1993) and the future (Altshuler 2016). Based on Altshuler & Stojnic (2016) and Stojnic & Altshuler (2017), I argue that the linguistic meaning of now — Kaplanean character, if you like — is not a time. Rather, now picks out the most prominent state, which holds throughout the time specified by the tense. But not any state will do. To satisfy the anaphoric constraints imposed by now, the state must be understood as resulting from a prominent event (possibly the speech event). In sum, I propose an analysis that advances the following two ideas: (i) state and time anaphora operate independently and (ii) the meaning of now relates these processes. I end the talk by: (a) showing how the proposed analysis sheds light on the oft-cited intuition that the use of now leads to a change-of-state inference (Recanati 2004) and (b) looking at constructions in which now takes a subordinate clause (Now that I’m in Chicago, I can see the White Sox play ), suggesting that now may be a dyadic connective on a par with while/when (Carter & Altshuler 2017).
Daniel Skibra: Epistemic-stative Sensitivity in Epistemic Modals and a Puzzle About Epistemic May
In this talk, I motivate the view that epistemic modals uniformly embed stative complements. I do so by considering two related issues. First, the aspectual dependence of patterns of modal interpretation, and second, some challenges to a prominent explanation of this puzzle. I argue that the challenges are met upon adopting this more uniform view. First, the pattern of aspectual dependency: certain modals, like must, clearly lack epistemic readings when they have eventive perfective complements. Epistemic readings are only attested with stative complements. (Derived stative constructions, like habituals and futurates also allow for epistemic readings.) So, (1) cannot have an epistemic reading, unless the predicate is interpreted as a habitual or a futurate. Following Ramchand (2014, ms.), let’s say that the pattern must exhibits here is one of epistemic-stative sensitivity (henceforth, ESS). It might seem, however, that not all modals exhibit ESS behavior. At first blush, modals like might seem not to exhibit epistemic-stative sensitivity; there is an acceptable epistemic reading of (2). The theoretical task, as Ramchand envisions it, is to explain what accounts for the behavior of ESS vs. non-ESS modals. Ramchand’s explanation appeals to two factors: 1) the role of clausal architecture in the composition of modal sentences, and 2) the distinction between so-called indexical and anaphoric modals, which differ in their ability to anchor certain types of anchoring situations. My talk will raise some skeptical remarks about the ability of the indexical/anaphoric distinction to adequately explain ESS behavior. An important challenge is brought by the behavior of the modal may. On Ramchand’s characterization, may is an indexical modal, but its putative ability to embed eventive prejacents patterns with anaphoric modals (like its near synonym might). Ramchand shrugs off the challenge by saying that the relevant readings of may are in fact root readings, and their interpretation as epistemic is merely apparent. I adduce some additional evidence to suggest that (3) is genuinely epistemic. I offer an alternative way to account for ESS behavior. Rather than shrugging off the epistemic reading of may, I challenge the view that the prejacent is in fact eventive. I offer these considerations as evidence that, surprisingly, eventive-stative sensitivity is in fact uniform for epistemic modals.
(1) John must go to the store.
(2) John might go to the store (later).
(3) John may go to the store (later).
Roger Schwarzschild: Stativity of Nouns
Questions to workshop: Simple nouns like horse are stative. What if we treat them the way we do stative verbs and adjectives, as predicates of states? How will they combine with quantifiers? How will DPs combine with verbs? What does plural marking on nouns signify? Is there anything to recommend this plan other than theoretical homogeneity?
Fabrizio Cariani and Alexis Wellwood: Comparative Confidence
The first part of the talk investigates the proper semantic treatment for confidence reports. Despite the recent surge of work on probability modals, there has been comparatively little work on gradable propositional attitudes like (1). We reject the idea that confident denotes measures of propositions (parameterized to the holder of the confidence). Our central argument rests on the possibility of nominal comparison, as in (2). Instead, we propose an analysis on which confident denotes properties of measurable states (again relativized to a subject). In the second part of the talk, we note several intuitive connections between graded confidence reports and ungraded belief reports in probabilistic claims: for instance, (1) sounds equivalent to (3). We use this observation to extend our states-based approach to probability operators, and contrast the resulting approach with other proposals by Yalcin, Lassiter and Klecha. If time permits, we will consider a further application to metalinguistic comparatives. (1) I’m more confident that she will ace syntax than semantics. (2) I have more confidence that she will ace syntax than semantics. (3) I think it is more likely that she will ace syntax than semantics.
Rebekah Baglini: Measuring States: Abstractness and Intensity
It is widely accepted that certain types of entities provide a natural basis for individuation and counting, while others do not. Entities may also vary in whether they are concrete or abstract spatiotemporally. In this talk, I argue that the core family of lexical stative meanings have a referential argument which admits neither counting nor measurement along an axis of physical extent. Through data drawn from English, Wolof (Niger Congo), and Amharic (Semitic), I demonstrate that this parametric approach provides the right foundation for capturing universal properties of stative predicates, despite variation in lexicalization strategies and morphosyntactic distribution.
Itamar Francez and Andrew Koontz-Garboden: Love and Happiness
We don’t know what love is, but in this talk we make a proposal for how to derive happiness. Specifically, building on the theory of qualities in our recent book (Francez and Koontz-Garboden 2017), we propose a compositional account of deadjectival nominalization like happiness and wisdom. We discuss whether this account can deal with various descriptive generalizations about the similarities and differences between English deadjectival nouns and other nouns, and whether it can meet some challenges posed to us by workshop participants Rebekah Baglini and Alexis Wellwood. On the account we propose, deadjectival nouns denote qualities: mutually exclusive sets of totally preordered and partially mereologically ordered entities. We consider the relation between qualities and other ontological categories that have been argued to play a role in the semantics of deadjectival nouns, such as states and tropes.
The common-sense notion of a state entails a lack of energy. Since we also have an intuition that energy is needed for causation, we should expect states not to be able to cause things to happen. Yet there is ample linguistic evidence suggesting that states do cause things to happen. In this talk I enumerate some of the ways states can be causal despite their lack of energy, and analyze a particular class of causal states in English have-causatives and futurates. The syntax-semantics interface guides us to an understanding of such states, which turn out to be at once transcendant and familiar. Throughout, the question of what is visible and what is invisible to the grammar is addressed.
Ashley Atkins: Back to the Futurate
Futurates have been described as sentences that convey that a future eventuality is planned, scheduled, or otherwise determined, but where this future orientation isn't obviously due to any overt linguistic element. Consider the futurate sentences, 'Mary is crossing the Atlantic tomorrow' and 'Mary crosses the Atlantic tomorrow,' for example. We appreciate that they both convey that a cross-Atlantic passage is due to take place tomorrow. But why aren't these sentences heard as contradictory? Why isn't the first heard as saying, for example, that a cross-Atlantic passage that occurs tomorrow is ongoing at the present time? 'Mary is crossing the Atlantic' is certainly heard as saying that a cross-Atlantic passage is presently ongoing. (Consider now the implausibility of assuming that 'tomorrow' contributes the meaning that such an event is due to take place the day after the day of utterance.) The central challenge posed by futurates is taken to be to account for the emergence of their future-oriented interpretations in the absence of any overt future-oriented meanings.
In her foundational work on futurate meaning, Bridget Copley pursued three assumptions that guided her toward a response to this challenge. The first is that these future-oriented interpretations are due to a covert modal meaning. The second is that this sort of meaning, though not uniform across various futurate expressions (e.g., progressive and simple) is linked to a system of meaning, and in particular, to the imperfective system, widely assumed to be a major natural language modal system. And third, any appeal to differences in event structure—the assumption that futurates are events with planning stages, for example—obscures rather than illuminates the nature of futurate meaning. I want to pursue a different course in my presentation, taking these three guiding assumptions as points of departure. I want to take seriously the possibility that the future orientation of futurates isn't due to any covert meaning (and so not to a covert modal meaning), that one of the important consequences of this assumption is that the imperfective system is not a modal system, and that appealing to structural differences—in particular, differences in the interactions between stative and telic meanings—can yield a uniform account of futurate meaning.
Daniel Altshuler: Distinguishing State and Time Anahora: A Look at Now
Since the work of Kaplan (1989) and Kamp (1971), the philosophical and linguistic tradition has treated now as a pure indexical: its linguistic meaning determines the referent on an occasion of use. In particular, its referent is the time of utterance on every occasion of use. There have been several challenges for this view, starting from the observation that now often occurs in narratives where it seems to pick out a time in the past (Dry 1979, Kamp & Rohrer 1983, Kamp & Reyle 1993) and the future (Altshuler 2016). Based on Altshuler & Stojnic (2016) and Stojnic & Altshuler (2017), I argue that the linguistic meaning of now — Kaplanean character, if you like — is not a time. Rather, now picks out the most prominent state, which holds throughout the time specified by the tense. But not any state will do. To satisfy the anaphoric constraints imposed by now, the state must be understood as resulting from a prominent event (possibly the speech event). In sum, I propose an analysis that advances the following two ideas: (i) state and time anaphora operate independently and (ii) the meaning of now relates these processes. I end the talk by: (a) showing how the proposed analysis sheds light on the oft-cited intuition that the use of now leads to a change-of-state inference (Recanati 2004) and (b) looking at constructions in which now takes a subordinate clause (Now that I’m in Chicago, I can see the White Sox play ), suggesting that now may be a dyadic connective on a par with while/when (Carter & Altshuler 2017).
Daniel Skibra: Epistemic-stative Sensitivity in Epistemic Modals and a Puzzle About Epistemic May
In this talk, I motivate the view that epistemic modals uniformly embed stative complements. I do so by considering two related issues. First, the aspectual dependence of patterns of modal interpretation, and second, some challenges to a prominent explanation of this puzzle. I argue that the challenges are met upon adopting this more uniform view. First, the pattern of aspectual dependency: certain modals, like must, clearly lack epistemic readings when they have eventive perfective complements. Epistemic readings are only attested with stative complements. (Derived stative constructions, like habituals and futurates also allow for epistemic readings.) So, (1) cannot have an epistemic reading, unless the predicate is interpreted as a habitual or a futurate. Following Ramchand (2014, ms.), let’s say that the pattern must exhibits here is one of epistemic-stative sensitivity (henceforth, ESS). It might seem, however, that not all modals exhibit ESS behavior. At first blush, modals like might seem not to exhibit epistemic-stative sensitivity; there is an acceptable epistemic reading of (2). The theoretical task, as Ramchand envisions it, is to explain what accounts for the behavior of ESS vs. non-ESS modals. Ramchand’s explanation appeals to two factors: 1) the role of clausal architecture in the composition of modal sentences, and 2) the distinction between so-called indexical and anaphoric modals, which differ in their ability to anchor certain types of anchoring situations. My talk will raise some skeptical remarks about the ability of the indexical/anaphoric distinction to adequately explain ESS behavior. An important challenge is brought by the behavior of the modal may. On Ramchand’s characterization, may is an indexical modal, but its putative ability to embed eventive prejacents patterns with anaphoric modals (like its near synonym might). Ramchand shrugs off the challenge by saying that the relevant readings of may are in fact root readings, and their interpretation as epistemic is merely apparent. I adduce some additional evidence to suggest that (3) is genuinely epistemic. I offer an alternative way to account for ESS behavior. Rather than shrugging off the epistemic reading of may, I challenge the view that the prejacent is in fact eventive. I offer these considerations as evidence that, surprisingly, eventive-stative sensitivity is in fact uniform for epistemic modals.
(1) John must go to the store.
(2) John might go to the store (later).
(3) John may go to the store (later).
Roger Schwarzschild: Stativity of Nouns
Questions to workshop: Simple nouns like horse are stative. What if we treat them the way we do stative verbs and adjectives, as predicates of states? How will they combine with quantifiers? How will DPs combine with verbs? What does plural marking on nouns signify? Is there anything to recommend this plan other than theoretical homogeneity?
Fabrizio Cariani and Alexis Wellwood: Comparative Confidence
The first part of the talk investigates the proper semantic treatment for confidence reports. Despite the recent surge of work on probability modals, there has been comparatively little work on gradable propositional attitudes like (1). We reject the idea that confident denotes measures of propositions (parameterized to the holder of the confidence). Our central argument rests on the possibility of nominal comparison, as in (2). Instead, we propose an analysis on which confident denotes properties of measurable states (again relativized to a subject). In the second part of the talk, we note several intuitive connections between graded confidence reports and ungraded belief reports in probabilistic claims: for instance, (1) sounds equivalent to (3). We use this observation to extend our states-based approach to probability operators, and contrast the resulting approach with other proposals by Yalcin, Lassiter and Klecha. If time permits, we will consider a further application to metalinguistic comparatives. (1) I’m more confident that she will ace syntax than semantics. (2) I have more confidence that she will ace syntax than semantics. (3) I think it is more likely that she will ace syntax than semantics.
Rebekah Baglini: Measuring States: Abstractness and Intensity
It is widely accepted that certain types of entities provide a natural basis for individuation and counting, while others do not. Entities may also vary in whether they are concrete or abstract spatiotemporally. In this talk, I argue that the core family of lexical stative meanings have a referential argument which admits neither counting nor measurement along an axis of physical extent. Through data drawn from English, Wolof (Niger Congo), and Amharic (Semitic), I demonstrate that this parametric approach provides the right foundation for capturing universal properties of stative predicates, despite variation in lexicalization strategies and morphosyntactic distribution.
Itamar Francez and Andrew Koontz-Garboden: Love and Happiness
We don’t know what love is, but in this talk we make a proposal for how to derive happiness. Specifically, building on the theory of qualities in our recent book (Francez and Koontz-Garboden 2017), we propose a compositional account of deadjectival nominalization like happiness and wisdom. We discuss whether this account can deal with various descriptive generalizations about the similarities and differences between English deadjectival nouns and other nouns, and whether it can meet some challenges posed to us by workshop participants Rebekah Baglini and Alexis Wellwood. On the account we propose, deadjectival nouns denote qualities: mutually exclusive sets of totally preordered and partially mereologically ordered entities. We consider the relation between qualities and other ontological categories that have been argued to play a role in the semantics of deadjectival nouns, such as states and tropes.