Events and the ontology of individuals: verbs as a source of individuating nouns

Events and the ontology of individuals: verbs as a source of individuating nouns

David Barner1, Laura Wagner2 & Jesse Snedeker1

1Harvard University 2 The Ohio State University

Address correspondence to: David Barner Laboratory for Developmental Studies

William James Hall, Harvard University 33 Kirkland Street Cambridge, MA 02138, USA

[email protected] Word count = 2982

Abstract We explored the mapping between mass-count syntax and semantics, and how verb event semantics affect the meaning of derived nouns. In Experiment 1, we compared the interpretation of mass and count nouns derived from iterative verbs that describe bounded events (e.g., jump) and nouns derived from non-iterative verbs that describe unbounded events (e.g., dance). Participants based judgments primarily on number for count nouns derived from both kinds of verb, and for mass nouns derived from iterative verbs (e.g., some jumping). In Experiment 2, we elicited dimension ratings that confirmed that number was the preferred dimension of comparison for all count nouns and for mass nouns derived from iterative verbs. We argue that these results provide evidence against the idea that mass nouns force an unindividuated construal (e.g., Bloom, 1994). We also suggest that verb iterativity is importantly linked to individuation by mass and count nouns, and may even access the same underlying ontology of individuals.

Keywords: individuation, syntax, semantics, boundedness, iterative events, mass nouns, count nouns

1. Introduction For the purposes of language, all “things” are equal. Noun phrases can refer to a

broad assortment of individual things, including objects, events, ideas, and emotions. These phenomena differ radically in nature, but natural language treats them uniformly in one respect: when used in count syntax, their names denote countable individuals. All words used in count syntax denote individuals (e.g., two books), while most mass nouns do not (e.g., some water). Also, particular words shift meaning when moved from mass syntax – where they quantify by mass or volume, for example – to count syntax – where they quantify by number (e.g., some string vs. some strings).1 This effect of syntax on construal led Quine (1960) to believe that count syntax itself is responsible for dividing experience into countable units, and has led others since to propose that only nouns used in count syntax can denote individuals (Link, 1983; Gordon, 1985; Landman, 1991; Bloom, 1994; Wisniewski, Imai & Casey, 1996). Such proposals raise the question of whether the linguistic ontology of individuals is indeed due to count syntax, or whether other syntactic structures (e.g., mass nouns or verbs) also permit individuation.

Two lines of enquiry suggest that count syntax is not required for individuation. First, several researchers have argued that words used in mass syntax can denote individuals (e.g., furniture; see Gillon, 1996; Chierchia, 1998). This intuition has been supported experimentally, both in adult speakers of English and 3- and 4-year-old children (Barner & Snedeker, 2005, under review; see also Bale & Barner, under review). Using the quantity judgment method, Barner and Snedeker (2005) demonstrated that adults and 4-year-olds judge six tiny shoes to be more shoes than two giant ones, but two large portions of mustard to be more mustard than six tiny portions. Crucially, participants also base quantity judgments on number for mass nouns like furniture to the same extent that they do for count nouns. Though count syntax forces individuation, mass syntax also permits reference to individuals when specified by particular words.

Second, several researchers have argued that names for things share an underlying ontology with event descriptions (whether nouns or verbs), such that both permit a grammatical encoding of experience as either “bounded” or “unbounded” (Gruber, 1967; Bach, 1986; Jackendoff, 1991; Talmy, 1985). The defining property of bounded events is that they are restricted by specific boundaries – spatial, temporal or abstract in nature. Because of this, when bounded events are described in the context of some interval or space that exceeds their boundaries, they “iterate”. For example, a punctual or “semelfactive” verb like jump (see Smith, 1991) cannot name a single event that extends over the course of a day, but must name a plurality of individual events (see example 1a). In contrast, a non-punctual verb like sleep can name a single protracted process that extends over a day without iteration (see example 1b). The important point is that certain verbs can encode reference to individuals, which requires an iterative interpretation when they are used in expressions that distribute phenomena over time.2

1 Count nouns, but not mass nouns, can appear in the singular plural, with cardinal numbers or quasi-cardinal determiners like these. Mass nouns can occur with terms like much. 2 We gloss over the link between boundedness and the telicity of verbs, since it is not clear that all verbs that name individuals are in fact telic (e.g., semelfactive verbs).

(1) a. Jim jumped all day (Jim performed multiple jumps) b. Jim slept all day (Jim slept continuously)

There is some evidence that iterative verbs and individuating nouns have common

underlying representations. Harley (1999) observes that whenever verbs are syntactically derived from individuating nouns, they name bounded individual events. For example, the noun “foal”, which denotes an individual, can be converted to a verb describing a bounded, iterative event (see 2a). However, the noun “drool,” which denotes an unbounded phenomenon, results in an unbounded, non-iterative, verb (see 2b; see Harley, 1999, for additional examples): (2) a. The mare foaled all day. (The mare had multiple foals)

b. Mary drooled all day. (Mary drooled continuously)

These examples suggest that certain words maintain the property of individuating or not individuating through their various syntactic incarnations, regardless of whether they denote objects or events. This is consistent with the claim that the ontology of linguistic individuals is not merely a product of particular syntactic constructions like count syntax, but is also a lexico-conceptual property of words which can be exploited by count syntax (see Barner & Snedeker, 2005, for details).

The evidence discussed thus far raises the question of whether individuation by mass nouns and verbs is due to independent phenomena with distinct explanations, or whether the same representations permit individuation in absence of count syntax in each case. If the individuation of mass nouns and verbs is due to a common lexico-conceptual feature, then mass nouns derived from verbs should inherit their individuating features. For example, if the conceptual features of iterative verbs like jump are inherited by mass nouns, then these derived nouns (e.g., some jumping) should denote individuals. Furthermore, these mass nouns should contrast with mass nouns derived from non-iterative verbs like dance (i.e. some dancing), which should quantify by some other dimension (e.g., time spent performing the action).3 Meanwhile, all count nouns derived from verbs should quantify by number, regardless of whether the original verb is iterative or non-iterative.

Evidence that the event semantics of verbs does affect the interpretation of derived mass nouns would be important for two reasons. First, it would provide the first experimental evidence for a link between the ontology of individuals for mass-count syntax and verb event semantics (i.e. iterativity). Second, evidence that iterative verbs lead to individuating mass nouns would provide additional evidence that mass syntax does not force an unindividuated construal. Challenges to this claim have been a topic of debate in linguistics and psychology for decades (Quine, 1960; McCawley, 1975; Link, 1983; Gordon, 1985; Gillon, 1996; Bloom, 1994; Chierchia, 1998; Barner & Snedeker, 2005), but have focused mainly on mass nouns that denote objects (e.g., furniture). The relative rarity of these words has led some to question whether they might be exceptions to an otherwise systematic mapping between mass-count syntax and individuation,

3 Despite sharing morphology with progressive verbs, gerundival mass nouns pass all of the standard distributional tests for being nouns.

mediated by their status as superordinate nouns. Thus, evidence from a productive category of words, like action-denoting mass nouns, is important for clarifying the actual interpretation of mass syntax. 2. Experiment 1

To determine how verb event semantics affects the meaning of derived mass nouns, Experiment 1 compared the interpretation of mass and count nouns derived from iterative or “semelfactive” verbs (see Smith, 1991) that name bounded events (e.g., jump) and nouns derived from non-iterative verbs that do not name bounded events (e.g., dance). 2.1. Methods

Participants were 64 English-speaking students recruited from the Harvard University campus. Participants were assigned to one of four conditions. Each condition included four short vignettes that described two characters performing actions. In two of the conditions the actions were “iterative” (jump, bet, kick, step), and in the remaining two they were “non-iterative” (dance, swim, smile, walk). Conditions presented the action words either in count syntax (e.g., some jumps), or in mass syntax (e.g., some jumping). Thus, iterativity and mass-count syntax were crossed between subjects. After each vignette, the participant was prompted to judge which of the two characters had performed more of the relevant action (henceforth called “quantity judgments”). Crucially, one character always performed a greater number of actions than the other, but his actions amounted to less overall time, distance or size. than the actions of the other character. This allowed us to test whether syntax led participants to base judgments on number or the other measuring dimension. Thus, the task was identical in structure to the quantity judgment method previously used to test mass-count quantification for solid objects (see Barner & Snedeker, 2005). A typical vignette is shown in Figure 1:

----- Insert Figure 1 about here ----- 2.2. Results and Discussion

Participants exhibited a clear asymmetry in their quantity judgments. They based almost all judgments on number for items used in count syntax, regardless of iterativity, but shifted their judgments for items used in mass syntax according to their iterativity (see Figure 2). Data were submitted to an ANOVA with two between-subjects factors: syntax (mass vs. count) and iterativity (iterative vs. non-iterative). There was a highly significant main effect of syntax, F (1, 60) = 106.30, p < .001, a main effect of iterativity, F (1, 60) = 50.24, p < .001, and crucially a significant interaction between mass-count syntax and iterativity, F (1, 60) = 37.47, p < .001. Planned comparisons indicated that participants based judgments significantly more on number for iterative mass nouns (75% of the time on average) than for non-iterative mass nouns (11%), t (30) = 7.51, p < .001, but revealed no significant difference between iterative (95%) and non-iterative (91%) items in count syntax, t (30) = 1.02, p > .3. Also, there was a significant difference between mass and count judgments for both non-iterative, t (30) = 2.55, p < .05, and iterative items, t (30) = 14.44, p < .001.

----- Insert Figure 2 about here ----- From these results we can conclude that count syntax imposes individuation, while

mass syntax leaves the specification of individuation up to particular lexical items. Thus, as suggested by previous studies, mass-count syntax exhibits asymmetry in its

interpretation (Barner & Snedeker, under review). Further, we can conclude that lexical properties that surface when words are used as verbs (e.g., iterativity) also play a role in determining the interpretation of corresponding derived mass noun usages. If a verb names an iterative event, its corresponding mass noun individuates; if it names a non-iterative event, the corresponding mass noun does not individuate. 3. Experiment 2

The first experiment used quantity judgments to test the interpretation of derived mass and count nouns. However, one possible problem with this method is that it assumes that the two dimensions being contrasted on any given judgment are those dimensions that participants would normally consider when quantifying the phenomena in question. For example, were we to ask participants whether jumping is better measured by number or intensity, they might select number despite ultimately preferring some third dimension, such as time or distance. To address this, Experiment 2 sought to replicate the results of Experiment 1 using a dimension-rating task, which asked participants to rank a large number of dimensions according to their relevance to judgments of amount. 3.1. Methods

Participants were 48 English-speaking students recruited from the Harvard University campus. Participants were shown one of four question sets. Each set included a page of instructions with an example of how ratings might be performed for two familiar nouns (friend and milk), plus four action words to rate. For each word, participants were asked to decide which of 10 dimensions were relevant in deciding who has done more of the relevant action, and then to rank these selections in order of their relevance. Participants were asked to rate either the iterative or non-iterative words from Experiment 1, either in mass or count syntax. Thus, iterativity and mass-count syntax were again crossed, between subjects. A typical item is shown in Figure 3.

----- Insert Figure 3 about here ----- 3.2. Results and Discussion

The results of Experiment 2 confirm that number is an important dimension for performing judgments of amount for all items used in count syntax, and iterative items used in mass syntax (see Figure 4). An ANOVA parallel to that in Experiment 1 was carried out, with the average number of times that number was chosen (out of 4 per subject) as the dependent variable. There was a significant main effect of syntax, F (1, 44) = 6.82, p < .05, a main effect of iterativity, F (1, 44) = 11.28, p < .005, and a significant interaction between syntax and iterativity, F (1, 44) = 6.82, p < .05. Planned comparisons confirmed that, for non-iterative items, number was mentioned more when items were presented in count syntax (96%) than in mass syntax (67%), t (22) = 2.61, p < .05. However, there was no difference in the selection of number as relevant between iterative count nouns (100%) and mass nouns (100%). Finally, number was mentioned significantly more often for iterative mass nouns than for non-iterative mass nouns, t (22) = 3.22, p < .005, though there was no difference between iterative and non-iterative count nouns, t (22) = 1.00, p > .3.

----- Insert Figure 4 about here ----- To determine the effect of iterativity and mass-count syntax on participants’

perception of how important number was for judging amount, we performed an ANOVA

comparing the number of trials (out of 4 per participant) in which participants ranked number as the most important dimension (see Figure 5). There was a significant main effect of syntax, F (1, 44) = 41.82, p < .001, a main effect of iterativity, F (1, 44) = 13.80, p < .005, and a significant interaction between syntax and iterativity, F (1, 44) = 4.26, p < .05. Planned comparisons found that participants ranked number as first in importance no more for iterative count nouns (98% of the time) than for non-iterative count nouns (86%), t (22) = 1.83, p > .05, though they ranked number first more often for iterative mass nouns (65% of the time) compared to non-iterative mass nouns (21%), t (22) = 3.24, p < .005. Finally, there was a significant effect of mass-count syntax for both non-interative nouns, t (22) = 6.25, p < .001, and iterative nouns, t (22) = 3.01, p < .01.

----- Insert Figure 5 about here ----- These results indicate that number is an important dimension of measurement for

iterative mass nouns and all count nouns, but not for non-iterative mass nouns. However, it is also apparent that number is more rigidly specified as the basis of quantification for count nouns than for iterative mass nouns, suggesting that mass syntax leaves judgments open to a broader array of measuring dimensions. To examine this, we performed a final analysis with the average number of dimensions mentioned per item (per subject) as the dependent variable, for iterative and non-iterative mass and count nouns. Overall, participants made reference to 2.88 dimensions on average for nouns in mass syntax (2.83 for non-iterative and 2.92 for iterative items), and 2.07 dimensions for nouns used in count syntax (2.06 for non-iterative and 2.08 for iterative items), resulting in a main effect of mass-count syntax, F (1, 44) = 11.51, p < .005, but no main effect of iterativity, F (1, 44) = 0.05, p > .8, and no interaction between syntax and iterativity, F (1, 44) = 0.02, p > .8.

----- Insert Figure 6 about here ----- Thus, participants judged that a broader array of dimensions were relevant to

performing judgments of amount when items were presented in mass syntax, compared to when the same items were presented in count syntax. This supports the idea that while count syntax specifies a uniform dimension of measurement (i.e. number), mass syntax is unspecified for measuring dimension.

4. General Discussion

This study of mass-count quantification provides the first experimental evidence for a link between verb event individuation and mass-count quantification. The lexical properties of individual verbs – i.e. their encoding of iterative events – are directly reflected in the interpretation of derived mass nouns. Mass nouns derived from iterative verbs lead speakers to base quantity judgments on number more frequently than those derived from non-iterative verbs. This effect of iterativity on mass nouns was also reflected by the frequency with which number was mentioned as relevant to measurement, and the frequency that number was ranked as first in importance. These results support the idea that iterativity in verbs and individuation by nouns result from related if not identical ontologies of individuals.

These results also provide strong evidence that mass syntax does not force an unindividuated construal. As noted above, various linguists and psychologists have proposed that only count nouns denote individuals and sets of individuals. Supporters of this view have suggested that apparent counterexamples – e.g., mass nouns like furniture

that name discrete, countable objects in the world – are unproblematic insofar as they fail to pick out individuals in the world qua individuals (see Bloom, 1994; McCawley, 1975; Wisniewski, Imai & Casey, 1996). Tasks that reveal judgments based on number may be based on the complexity of physical objects in an array (rather than their linguistic representations) or on the semantics of subordinate noun forms (e.g., chair in place of furniture).

However, the results of this study support a different conclusion. First, it is true that count syntax imposes quantification over individuals, even if corresponding verbs or mass nouns do not. This is demonstrated here by event-denoting words like dance and swim. Though the relevant mass usages do not quantify over individuals, the count usages do. Second, mass nouns derived from iterative action verbs also denote individuals. This is demonstrated by event-denoting words like jump. These words do not denote complex physical objects, nor are they superordinate terms that might be subject to special semantic considerations. Instead, they are abstract terms that provide linguistic criteria for individuation, just like count nouns. Third, speakers select a broader array of measuring dimensions as relevant to interpreting mass nouns compared to count nouns, for which number is of primary importance. Based on these results, we conclude that there is a fundamental asymmetry between mass and count syntax: count syntax specifies a uniform measuring dimension (i.e. number), while mass syntax leaves the dimension of measurement open to specific lexical items.

To summarize, we have presented evidence that the linguistic ontology of individuals employed by mass and count nouns is closely related if not identical to that encoded by verbs. Just as all things are equal for the purposes of language, so too are many components of language equal for the purposes of individuation. Acknowledgements This research was supported by a fellowship from the National Sciences and Engineering Research Council of Canada to D.B.

References Bach, E. (1986). The algebra of events. Linguistics and Philosophy, 9, 5-16. Bale, A., & Barner, D. (under review). Comparative dimensions and individuation: The

semantics of mass and count nouns. Barner, D., & Snedeker, J. (2005). Quantity judgments and individuation: Evidence that

mass nouns count. Cognition, 97, 41-46. Barner, D., & Snedeker, J. (under review). Children's early understanding of mass-count

syntax: individuation, lexical content, and the number asymmetry hypothesis. Bloom, P. (1994). Semantic competence as an explanation for some transitions in

language development. In Y. Levy (Ed.), Other children, other languages: Theoretical issues in language development. Hillsdale, NJ: Erlbaum.

Chierchia, G. (1998). Plurality of mass nouns and the notion of "semantic parameter". Events and Grammar, 70, 53-103.

Gillon, B. (1996, June). The lexical semantics of English count and mass nouns. Paper presented at the Workshop on the Breadth and Depth of Semantic Lexicons, Santa Cruz.

Gordon, P. (1985). Evaluating the semantic categories hypothesis: the case of the mass/count distinction. Cognition, 20, 209–242.

Gruber, J. (1967). Functions of the lexicon in formal descriptive grammar. Santa Monica: Systems Development Corporation. Reprinted as part of Lexical structures in Syntax and Semantics, 1976. North Holland Amsterdam.

Harley, H. (1999). Denominal verbs and aktionsart. In L. Pylkanen and A. van Hout, Eds., Proceedings of the 2nd Penn/MIT Roundtable on Event Structure, MITWPL: Cambridge.

Jackendoff, R. (1991). Parts and boundaries. Cognition, 41, pp. 9-45. Landman, F. (1991). Structures for semantics. Dordrecht: Kluwer Academic Publishers. Link, G. (1983). The logical analysis of plurals and mass terms: A lattice-theoretical

approach. In Bauerle, R., Schwarze, C., Stechow, A. (Eds.), Meaning, use, and interpretation of language. Berlin: de Gruyter.

McCawley, J. (1975). Lexicography and the count-mass distinction. Proceedings of the First Berkeley Linguistic Society Conference, 314-321.

Smith, C.S. (1991). The parameter of aspect. Dordrecht, Kluwer. Talmy, L. (1985). Lexicalization patterns: semantic structure in lexical forms. In T.

Shopen (Ed.) Language Typology and Syntactic Description III: Grammatical Categories and the Lexicon, Cambridge University Press, Cambridge.

Quine, W.V.O. (1960). Word and object. Cambridge, MA: MIT Press. Wisniewski, E.J., Imai, M., & Casey, L. (1996). On the equivalence of superordinate

concepts. Cognition, 60, 269-298.

Jerry and Jake like to swim in the lake. Recently they swam the following distances: Jerry swam two distances Monday: Once back and forth across the lake (250 meters) Wednesday: Twice back and forth across the lack (500 meters) Jake swam three distances Monday: To the dock and back (20 meters) Tuesday: Twice to the dock and back (40 meters) Wednesday: To the dock and back (20 meters) Question: Overall, who did more swimming? Fig. 1. Example vignette and quantity judgment from Experiment 1, using mass syntax and the atelic mass noun “swimming”

Fig. 2. Quantity judgments for iterative and non-iterative mass and count nouns

Who did more … SWIMMING?

Fig. 3. Example of a dimension rating item from Experiment 2, for the mass noun derived from the unbounded verb “swim”

Fig. 4. Percentage of trials that number was selected as relevant to judging quantity for iterative and non-iterative mass and count nouns

Can be used to calculate amount?

Dimension Rank

Diameter Volume Height Number Mass Time Value Intensity Distance Depth

Fig. 5. Percentage of trials that number was ranked first in importance for judging quantity for iterative and non-iterative mass and count nouns

Fig. 6. Average number of dimensions selected as relevant for judging quantity for iterative and non-iterative mass and count nouns