Perspectives on Psychological Science

Trading Spaces: Carving Up Events for Learning 5(1) 33-42

ª The Author(s) 2010
Reprints and permission: http://www.
Language sagepub.com/journalsPermissions.nav
DOI: 10.1177/1745691609356783
http://pps.sagepub.com

Tilbe Göksun1, Kathy Hirsh-Pasek1, and

Roberta Michnick Golinkoff 2
1
Department of Psychology, Temple University, Philadelphia, PA and 2School of Education and Departments of Psychology
and Linguistics and Cognitive Science, University of Delaware, Newark

Abstract
Relational terms (e.g., verbs and prepositions) are the cornerstone of language development, bringing together two distinct fields:
linguistic theory and infants’ event processing. To acquire relational terms such as run, walk, in, and on, infants must first perceive
and conceptualize components of dynamic events such as containment–support, path–manner, source–goal, and figure–ground.
Infants must then uncover how the particular language they are learning encodes these constructs. This review addresses the
interaction of language learning with infants’ conceptualization of these nonlinguistic spatial event components. We present
the thesis that infants start with language-general nonlinguistic constructs that are gradually refined and tuned to the requirements
of their native language. In effect, infants are trading spaces, maintaining their sensitivity to some relational distinctions while dam-
pening other distinctions, depending on how their native language expresses these constructs.

Keywords
language development, relational terms, event processing, foundational event components

If we are ultimately to understand how children learn to express review considers what we have learned thus far. The issue
the semantics of their language, we will need to understand the we address is how learning a language interacts with the con-
conceptual foundations on which those semantics rest. (Man- ceptualization of four nonlinguistic foundational constructs:
dler, 2004, p. 281) containment–support, path–manner, source–goal, and figure–
ground. Reviewing evidence from four distinct domains, this
Learning relational terms such as verbs and prepositions is article offers new avenues for both theoretical and practical
fundamental to language development. Verbs, in particular, are research.
centerpieces of sentences. Verbs and prepositions afford us the
capability to describe static and dynamic relations between
objects and participants in events (e.g., the cup is on the table What Does It Take to Learn Relational
or the dog is chasing the woman). In some languages (e.g.,
Terms?
Korean), verbs are used to capture the spatial relations that
English reserves for its prepositions. For example, whereas Relational term learning is a two-step process. First, infants
English uses a single preposition for on (as in ‘‘put a cap on must perceive the actions and events that languages express.
a pen’’ or ‘‘put an apple on one’s hat’’), Korean has many verbs Second, infants must learn which event components are
for on, depending on what article of clothing is being put on encoded in their native language and how their language
and how tightly the on relation appears (as a cap on a pen packages these components (Gentner, 1982; Gentner &
versus an apple on a table). The study of the acquisition of rela- Boroditsky, 2001; Gentner & Bowerman, 2009; Golinkoff &
tional terms draws on literature from two distinct fields: lin- Hirsh-Pasek, 2008; Tomasello, 1995). To do this, infants must
guistic theory and infants’ event processing. Research has perceive and individuate the actions within events, categorize
begun to investigate the acquisition of relational terms (e.g.,
Casasola & Cohen, 2002; Choi & Bowerman, 1991; Golinkoff
Corresponding Author:
& Hirsh-Pasek, 2008; Lakusta, Wagner, O’Hearn, & Landau, Tilbe Göksun, Department of Psychology, Temple University, Temple Infant
2007; Mandler, 2004; Pulverman, Golinkoff, Hirsh-Pasek, & Lab, 580 Meeting House, 1st Floor of Haines House, Ambler, PA 19002
Sootsman-Buresh, 2008; Shipley & Zacks, 2008), and this E-mail: [email protected]

33
34 Göksun et al.

these actions, and learn how to map words onto these actions These theories share the belief that children analyze the
(Golinkoff et al., 2002). Given that languages comment on dif- events taking place around them and learn to focus on just those
ferent aspects of the same event, this is a daunting task. For aspects that their language expresses. They differ in other
example, to learn the verb march, an English-reared infant respects: Whereas Slobin is agnostic to the source of the non-
must differentiate the act of marching from, say, jumping. The linguistic constructs language encodes, Mandler considers
Turkish-reared infant must make this distinction as well. these constructs to be conceptual primitives, available very
However, the act of marching is encoded as a verb in English early in development. Here we present the thesis that infants
(‘‘march into the class’’), whereas in Turkish (‘‘sınıfa start with language-general concepts that are gradually con-
yürüyerek girdi—go into the class marchingly’’), it surfaces strued in language-specific ways. Sensitivity to distinctions in
as an adverb, if at all. In her natural partitions hypothesis, events becomes refined or abandoned as the conceptual frame-
Gentner (1982) claimed that ‘‘lexicalizing’’ relational terms work makes contact with language. Therefore, infants bring
is more demanding than simply perceiving movement, connec- new perspectives to their interpretation of spatial and event
tions between actors, and directional changes within events components. In a way then, they are ‘‘trading spaces’’ as they
(Gentner & Bowerman, 2009). Tomasello (1995) called this the learn language.
‘‘packaging problem’’: The child must discern to which aspect Consider an imperfect analogy from the domain of phonolo-
of an event an adult is referring. gical development. At the start of language learning, infants
around the world possess an auditory system that affords them
the ability to distinguish between phonemes in the world’s lan-
guages, regardless of the language to which they are exposed
Foundational Constructs in Events (e.g., Eimas, Miller, & Jusczyk, 1987; Kuhl et al., 1997;
Talmy (1985) outlined a number of components that describe Werker & Lalonde, 1988; Werker & Tees, 1984). However,
the relational terms codified across languages (see also exposure to the particular phonological contrasts of their native
Jackendoff, 1983; Lakoff, 1987; Langacker, 1987; Talmy, tongue apparently lessens the ability to make phonological dis-
2000). Among them are path, or the trajectory of an action with tinctions that appear in nonnative languages. Thus, perceptual
respect to a ground (e.g., over or under); manner, or how the reorganization (e.g., Galles-Sebastian, 2006; Kuhl, 2004;
action is performed (e.g., jumping or rolling); figure, (the mov- Werker & Tees, 1984) occurs when infants narrow the spec-
ing or conceptually movable entity) and its relation to the trum of sounds that they attend to between those encoded by
ground (the reference entity or a stationary setting); and source their native language and those not encoded.
(beginning point of an event) and goal (ending point of an Infants might learn relational language in a similar fashion
event). Other constructs refer to spatial relations (Choi & (for similar arguments see also Choi, 2006; Clark, 2003,
Bowerman, 1991; Talmy, 1985) like containment (putting 2004; Hespos & Spelke, 2007). They might notice a common
things in a container) and support (putting things on a surface). set of foundational components of events regardless of the lan-
Conceptual foundations such as these create the semantic bases guage they are learning. Then, influenced by distinctions
for world-to-word relations. encoded in the native language, they might focus on a subset
This linguistic taxonomy for relational terms meets psycho- of these components that are relevant to their native language.
logical theory in two dominant theories. Slobin (1996) suggests Analogously, this phenomenon might be called semantic
that languages are not ‘‘neutral coding systems of an objective reorganization, in which universal perceptual constructs are
reality’’ (p. 88). That is, the very same event will be described reorganized to match the expressional tendencies of one’s
differently depending on the language. To learn to think for native tongue. Language, in this case, would have the function
speaking (Slobin, 1996, 2001), children must notice the set of of orienting infants’ attention to some relations in events over
distinctions that speakers make in daily conversation. Thus, others.
language learners pay attention to events and to how their To evaluate this thesis, we revisit some of the key semantic
particular language community encodes aspects of those events distinctions proposed by linguists (Jackendoff, 1983; Talmy,
in the ambient language. 1985) that are available in events and lexicalized across
Mandler (1992, 2004) adopts an explicitly developmental languages differently by relational terms. By examining
perspective focusing on how children view the events that lan- semantic reorganization across four domains, this article offers
guage will encode. She suggests that prior to language, infants a unique panoramic view of the interaction between infants’
construct image schemas to store fundamental meanings that nonverbal conceptual processing of nonlinguistic event compo-
derive from perceptual meaning analysis (i.e., through atten- nents and their expression in language.
tion, infants redescribe perceptual information into a simpler
form that reaches awareness). Common image schemas are
Processing Nonlinguistic Foundational
those of path, link, containment, and support, which are later
combined to derive basic conceptual categories such as ani-
Constructs
macy, causality, and agency. For example, by noting a figure’s An event can be defined as ‘‘a segment of time at a given loca-
ability to rapidly change path (without apparent external tion that is perceived by an observer to have a beginning and an
impetus), infants come to identify animate objects. end and their relations’’ (Zacks & Tversky, 2001, p. 3). Before

34
Trading Spaces 35

infants process components of events like path–manner or the objects (i.e., degree of fit). The spatial verb kkita, which
source–goal, we need to ask whether infants use events as crosscuts the English categories of put in and put on, describes
psychological units. a tight-fitting relation between the objects (Choi & Bowerman,
Research suggests that infants discriminate changes in pat- 1991; Gentner & Bowerman, 2009). Putting a ring on a finger
terns of motion (e.g., Bogartz, Shinskey, & Schilling, 2000; and putting a book in a cover are both described with the verb
Cashon & Cohen, 2000) and remember specific patterns kkita in Korean (Choi, 2006).
(Bahrick & Pickens, 1995). During the first year, infants can Using both looking time and reaching behavior as dependent
distinguish biological motion from nonbiological motion for variables, Baillargeon and her colleagues show that infants are
both people and other mammals (Arterberry & Bornstein, capable of discriminating the spatial relations of containment,
2002; Bertenthal, 1993), identify both rational and intentional support, occlusion, and covering by 6 months of age (e.g.,
actions (Csibra, Gergely, Biro, Koos, & Brockbank, 1999; Aguiar & Baillargeon, 1999; Baillargeon, 2004; Baillargeon,
Woodward, 1999), and reason about the physical interaction Needham, & DeVos, 1992; Baillargeon & Wang, 2002; Hespos
between objects such as causality (e.g., Leslie, 1982; Oakes, & Baillargeon, 2001a, 2001b, 2008; Hespos & Piccin, 2009).
1994). Infants also parse actions in events (e.g., Baldwin, Further, even English-reared 5-month-olds distinguish between
Baird, Saylor, & Clark, 2001; Sharon & Wynn, 1998; Spelke, tight-fit and loose-fit events in both containment and support
Born, & Chu, 1983; Wynn, 1996). Once infants attend to and categories (Hespos & Spelke, 2004), demonstrating that prelin-
represent events, they must also detect those aspects of events guistic infants are sensitive to spatial distinctions that are not
that are related to linguistic expressions (Clark, 2003). lexicalized in their native language. Hespos and Piccin
To make the case that infants are sensitive to event (2009) also demonstrated similar patterns in covering events.
constructs that will be realized differently across various lan- Six-month-old infants categorized containment relations
guages, we need to illustrate how infants detect the specific dis- (Casasola, Cohen, & Chiarello, 2003), but support relations
tinctions of event components realized in the worlds’ languages were not categorized before 14 months of age unless the task
and show that they can categorize these components (Golinkoff was simplified, as when the number of exemplars of the
& Hirsh-Pasek, 2008). We do not mean to imply that these are category was reduced (Casasola, 2005b). In addition, both
the only conceptual distinctions infants attend to when reason- English- and Korean-reared 9-month-old infants categorized
ing about events—rather, we are focusing our work only on events observing the common degree-of-fit relation (i.e., tight-
those constructs that are central to language processing. or loose-fit; McDonough, Choi, & Mandler, 2003), considering
Four event components closely examined in the literature ‘‘a key in a keyhole’’ to be the same relation as ‘‘a cork in a
are containment–support, path–manner, source–goal, and bottle.’’
figure–ground. These constructs share three features. First,
they are perceptually accessible to infants (Mandler, 2004). For
these constructs to be useful for language, they must be noticed
Path–Manner
and categorized across different actors and locations. Second, Path is defined as a figure’s trajectory relative to a ground, and
these components are universally codified across languages manner refers to how the action is performed. For example, in
(Jackendoff, 1983; Talmy, 1985, 2000). For example, the path the sentence ‘‘John is running into the room,’’ John is the fig-
of an event is expressed in many languages with verbs (e.g., ure, running is the manner, and into is the path of the event.
descend, exit) and prepositions such as into and across. Third, However, English often conflates motion with manner in the
although they are all linguistically expressed, languages differ main verb (as in running) and expresses the path in a ‘‘satel-
in the ways in which they encode these constructs (e.g., English lite’’ prepositional phrase (‘‘ . . . into the room’’). In contrast,
uses climb up, whereas Turkish uses tırmanarak çıktı ‘‘go up Turkish conflates the motion with path in the main verb (as
climbingly’’). in girdi: ‘‘go into’’) and expresses manner in a subordinated
The common features among these constructs, as well as the verb or adverbial phrase (kosarak: ‘‘runningly’’)
burgeoning empirical data in these four areas, allow us to dis- Seven-month-old English infants attend to path and manner
cuss our thesis in an integrated way. Thus, we will present changes in nonlinguistic dynamic events (Pulverman &
infants’ nonlinguistic conceptualization of these foundational Golinkoff, 2004). In particular, after being habituated to an ani-
constructs under four subheadings. mated starfish performing both a path and a manner (e.g., a
starfish twisting over a ball) in test trials, infants increased their
attention to both a path change (e.g., starfish twisting under a
Containment–Support ball) and a manner change (e.g., starfish spinning over a ball).
A containment relation occurs when something is fully or par- Similar results were obtained from Spanish-reared and
tially surrounded by a container (e.g., in), and a support relation Mandarin-reared infants (Pulverman, Chen, Chan, Tardif, &
refers to the contact of an object on top of surface (e.g., on). Meng, 2007; Pulverman et al., 2008).
Although many languages use terms similar to in and on in Infants also categorize paths and manners by 10 and
encoding containment and support, they express them in vastly 13 months of age, respectively, when these are performed in
different ways. In Korean, for example, containment and sup- an invariant manner. For example, upon seeing the same path
port events are labeled on the basis of tight or loose fit between (e.g., under) presented with different manners (e.g., spinning,

35
36 Göksun et al.

twisting, toe-touching; Pruden, Hirsh-Pasek, Maguire, & from linguistics to the study of event perception and language
Meyer, 2004), children notice when the path changes but not development.
when the manner changes in test events. Ten- to 15-month-
old infants also formed nonlinguistic categories of two manners
(i.e., hopping and marching) over five different actors (Song
et al., 2006).
The Role of Language
Once they isolate and categorize components in events, chil-
dren need to lexicalize these event components in their native
Source–Goal language. Language might assist toddlers attuning their con-
In a motion event, source refers to the figure’s movement from ceptual distinctions to their native language (Spelke & Hespos,
a reference object by a variety of possible ‘‘from or away from 2002). Here, we explore three issues related to the role lan-
paths.’’ Thus, a source might be the chair from which the dog guage plays in processing events: (a) how the vocabulary level
moves toward his bowl, which is the goal. Goal refers to the of the child relates to perception of the event components, (b)
figure’s movement to a reference object, using ‘‘to or towards how the presence of labels facilitates the abstraction of these
paths’’ (Jackendoff, 1983; Talmy, 1985). Languages code goals constructs from events, and (c) how language learning interacts
more frequently than sources, possibly because the endpoint of with the interpretation and expression of these components.
an event is more important for further action.
Twelve-month-olds prefer to attend to goals rather than
sources in nonlinguistic dynamic events (Lakusta et al., Vocabulary Knowledge
2007), corroborating the frequently reported goal bias in the lit- One might expect that children’s nonnative analysis of event
erature (e.g., Csibra & Gergely, 1998; Regier & Zheng, 2007; components would be inversely related to vocabulary level in
Woodward, 1998). Apparently, 14-month-old infants can form their native language. That is, we might hypothesize that
a goal category, but not a source category involving different children who have more words in their lexicons relative to their
goal objects, spatial relations, and agents (Lakusta & Carey, peers might be worse at noticing nonnative semantic distinc-
2008). tions, whereas children with fewer words might still differenti-
ate between event components not expressed in their native
language. This assumption is similar to the weak analogy from
Figure–Ground phonetic discrimination. For example, English-reared 7-month-
The figure in an event can follow any path or move from any olds who were better at discriminating native phonemes
source. The ground is a stationary setting with respect to a fig- produced a greater number of words and larger utterances with
ure’s movement. For example, in the sentence ‘‘John is walking greater sentence complexity at 14, 18, 24, and 30 months. In
across the street,’’ John is the figure and the street is the ground. contrast, better nonnative phoneme discrimination (Mandarin
Notably, figure and ground are packaged differently in lan- Chinese) reduced later language ability (Kuhl, Conboy,
guages like English and Japanese. Japanese ground–path verbs Padden, Nelson, & Pruitt, 2005; see also Tsao, Liu, & Kuhl,
such as wataru (‘‘go across’’) or koeru (‘‘go over’’) incorporate 2004).
constraints on the physical geometry of the ground along with Studies on event components confirm that vocabulary size
the direction of motion (Muehleisen & Imai, 1997). For exam- correlates with the detection of nonnative semantic distinc-
ple, wataru implies two things: (a) that there is both a starting tions. English-speaking 29-month-old children with more
point and a goal, and (b) that the ground should be a flat words in their vocabularies relative to their peers or the ability
extended surface. The typical grounds for wataru are railroad, to produce the word in were less likely to perceive the differ-
road, or bridge. In contrast, when the ground does not contain a ence in the Korean degree-of-fit than were low vocabulary
barrier between two sides (e.g., a tennis court, grassy field), the children or those who did not yet produce the word in (Choi,
verb tooru (‘‘go through’’) is used. 2006). In contrast, Korean-speaking children at the same age,
English-reared infants differentiate figures (e.g., a man or a regardless of vocabulary level, still demonstrated sensitivity
woman crossing a railroad) and grounds (e.g., crossing a rail- to tight-fit versus loose-fit containment categories. Likewise,
road vs. crossing a tennis court) in dynamic events by 10 and Pulverman et al. (2008) found that 14- to 17-month-old
13 months of age, respectively. It is important to note that the English-reared infants who had greater vocabularies by mater-
same infants distinguish grounds that are coded differently by nal report were more attentive to manner changes than to path
Japanese ground–path verbs (e.g., crossing a railroad vs. a changes, which mirrors English’s vastly greater number of
grassy field; Göksun, Hirsh-Pasek, & Golinkoff, 2008, 2009). manner than path verbs. On the other hand, Spanish-reared
In sum, empirical data across these four domains suggest infants with low vocabularies paid more attention to manner
that infants possess a set of nonlinguistic constructs that form than did their high-vocabulary counterparts. Spanish uses path
the bases for learning relational language. Infants appear to dis- verbs and has very few manner verbs. Perhaps the low-
criminate and form categories of these components of dynamic vocabulary Spanish learners were still attending to the event
events. These four lines of research suggest criteria for good component that is less frequently encoded in Spanish. Alterna-
candidates of foundational semantic constructs when we move tively, attention to manner might delay Spanish-reared infants’

36
Trading Spaces 37

ability to learn more verbs (Pulverman, Hirsh-Pasek, (McDonough et al., 2003; see also Hespos & Spelke, 2004). In
Golinkoff, Pruden, & Salkind, 2006). contrast, Munnich, Landau, and Dosher (2001) did not find dif-
These findings taken together suggest that learning language ferences in nonlinguistic tasks for contact–support relations
dampens the detection of categorical differences that are not when testing English-, Japanese-, and Korean-speaking adults
encoded in one’s native tongue. To the extent that vocabulary (see also Norbury, Waxman, & Song, 2008). The only difference
is a reflection of native language learning, children who acquire among language groups appeared when people named these
more words are more likely to make native distinctions in relations (but see Boroditsky, 2001; Boroditsky & Ramscar,
events and less likely to make nonnative distinctions. 2002). Nevertheless, the ability to note nonnative spatial rela-
tionships is not completely lost, as adults’ attention can be drawn
to note these distinctions (Hespos & Spelke, 2004).
Labeling a Target Event Component When do toddlers and preschoolers make language-specific
Prior research has shown that labeling increases attention to interpretations of event components? Few studies have as yet
objects (e.g., Baldwin & Markman, 1989) and facilitates cate- examined this question. Maguire and her colleagues found
gorization of both familiar and novel objects (e.g., Balaban & that English-, Spanish-, and Japanese-speaking 2.5-year-olds
Waxman, 1997; Booth & Waxman, 2002, 2003; Fulkerson & preferred to extend a novel verb to the path of the action, but
Haaf, 2003; Waxman, 1999; Waxman & Booth, 2003). Does 3-year-olds speaking these languages presented more language-
labeling promote or hinder the detection of components in specific patterns of verb construal. For example, English-
dynamic events? Pulverman, Golinkoff, Hirsh-Pasek, and Bran- speaking children assume that a novel verb labels manner, and
done (2009) presented 14- to 17-month-old English-reared Spanish-speaking children are less likely to interpret the novel
infants with the same videos used to test children in silence by verb as manner (Maguire et al., in press).
Pulverman et al. (2008) but used either a noun label (i.e., He’s This asymmetry in encoding nonlinguistic event compo-
a jame!) or a verb label (e.g., He’s jaming!) only during habitua- nents also appears in children’s and adults’ linguistic expressions
tion. Hearing a verb, children increased attention to manner but of events. Choi and her colleagues demonstrated that English-
not path in test trials, suggesting that a novel verb label selec- and Korean-speaking children use spatial terms for containment
tively influences infants’ event processing. In addition, a novel and support in language specific ways, starting at around 2 years
verb, but not a novel noun, enhanced attention to events. In a of age (Bowerman & Choi, 1994; Choi & Bowerman, 1991).
potential verb learning task with appropriate labels, English- Similarly, children encode language-specific patterns for path
reared infants increasingly attend to the most frequently and manner starting at 3 years of age (e.g., Allen et al., 2007;
expressed component of events in English: the manner of motion. Özçaliskan & Slobin, 1999; Papafragou, Hulbert, & Trueswell,
Does labeling also facilitate infants’ categorization of event 2008; Papafragou, Massey, & Gleitman, 2006). For example,
components? Casasola (2005a) found that hearing the familiar Papafragou et al. (2006) found that Greek-speaking children and
word on helped 18-month-old infants to abstract the category of adults mentioned the path of the motion significantly more than
support for both familiar and novel objects. Similarly, the use the manner, which is consistent with the dominance of path
of a novel verb label (e.g., javing) aids earlier categorization verbs in Greek, whereas English speakers demonstrated the
of paths and manners at 7 and 10 months of age, respectively opposite encoding. The cross-linguistic analyses on the expres-
(Pruden & Hirsh-Pasek, 2006). sion of source and goal indicate that both adults and children are
These findings suggest that both familiar and novel labels more likely to talk about endpoints than starting points in motion
buttress the detection and categorization of foundational event events (Johanson, Selimis, & Papafragou, 2008; Regier &
components. Yet the precise role of labeling in influencing the Zheng, 2007) and that typically developing and deaf children
formation of spatial event categories is still unclear. Labeling manifest a goal bias in their use of language and sign, respec-
might heighten the similarities between events. tively (Lakusta & Landau, 2005; Zheng & Goldin-Meadow,
2002). Thus, an astounding and ‘‘universal’’ goal bias is main-
tained in both nonlinguistic event processing and linguistic
Event Interpretation and Expression by Adults
descriptions presumably because languages code goals more fre-
and Children quently than sources (for Japanese findings, see Lakusta,
If language has an impact on which event components children Yoshida, Landau, & Smith, 2006).
attend to, perhaps it also influences how adults’ and toddlers’ In sum, at around 3 years of age, children become language-
perception of events in silence. This weak version of the Whor- specific event interpreters as they gain more experience with
fian hypothesis predicts that people should interpret nonlin- their native tongue. These findings suggest that children
guistic events differently depending on their native language. restructure the available nonlinguistic spatial constructs with
What does the research tell us about how adults from different respect to the language being learned.
linguistic environments interpret the same nonlinguistic
events? Does the language they speak influence their percep-
tion? Studies suggest that Korean-speaking adults, but not
Trading Spaces
English-speaking adults, differentiated between tight-fit and This article tracked infants’ nonverbal conceptual processing
loose-fit containment in a nonlinguistic discrimination task of nonlinguistic event components and how children learn

37
38 Göksun et al.

about the way in which these event components are expressed trajectories might reflect the perceptual saliency of some of the
in their native language. The literature suggests that infants components over others. Regier and Zheng (2007) suggested
detect and categorize at least four conceptual categories that attention might shape language such that elements of
described here by the beginning of the 2nd year of life. With events that universally attract attention might induce linguistic
these constructs in place, the underpinnings for the learning semantics. For example, the spatial configuration of a resulting
of a language’s relational terms are in place. As children lexi- event (goal) is more salient and accessible than a starting event
calize these components in their native tongue, they appear to (source). Thus, both nonlinguistic visual discriminations and
tune into certain semantic distinctions over others, influenced language encodings favor the endpoint of events across lan-
by the ambient language. Furthermore, there is the suggestion guages (Regier & Zheng, 2007). Just how much the perceptual
that the more language they know, the more attentive they are environment influences language or how language heightens
to native over the nonnative encodings of these constructs. attention to perceptual information is still hotly debated (e.g.,
Trading spaces occurs when a semantic component (such as Bowerman & Levinson, 2001; Li & Gleitman, 2002; Munnich
containment or support) is semantically reorganized to match et al., 2001).
the expression of that component in the ambient language. In The third conclusion is that, with development and
fact, the native language might play a causal role in how increased exposure to the ambient language, children begin
children divide their spatial world, as they gradually adopt the to package these nonlinguistic constructs in the way that they
particular relational terms their language uses. Unlike reorgani- are encoded in their native language (e.g., Allen et al., 2007;
zation in phonological development, however, reorganization Choi, 2006). Children seem to interpret events along the lines
in semantic development refers to the hierarchy of preferences of the statistical tendencies of their native language and assume
people develop and not to the loss of the ability to note these that speakers will package language in ways consistent with
nonnative event distinctions in the absence of lengthy training their language. As Li and her colleagues suggested ‘‘Speakers
as adults (Tees & Werker, 1984). will use differences in language patterns as a probabilistic basis
This article differs from previous discussions about the for inferring how new words and sentences will relate to new
similarities between phonological and semantic development objects and events . . . the words and sentences we utter map
that only hinge on the categories of containment–support only very approximately onto the thoughts we mean to express,
(Choi, 2006; Hespos & Spelke, 2004, 2007) as it adds force a truism that requires humans to apply considerable inferential
to the argument by extending it to the dynamic event compo- analysis to make sense of the speech of their interlocutor’’ (Li,
nents of path–manner, source–goal, and figure–ground. The Abarbanell, Gleitman, & Papafragou, 2009, p. 35). The orien-
view that semantic reorganization takes place in early develop- tation toward the native language’s distinctions and encoding
ment is systematically strengthened by the inclusion of other system—thinking for speaking—can only occur after sufficient
spatial event constructs. language is learned.
Our analyses yield three broad conclusions. The first is that We are not here arguing in favor of a position that endorses
infants come prepared to divide the events in their world into a Whorfian linguistic relativity (Whorf, 1956). That position
universal set of categories that are relevant to later language. proposes that the learned language affects the way people
They parse events and abstract these components in ways that think. Rather, it appears that language exposure increases sen-
lay the groundwork for the learning of relational terms like sitivity to some aspects of events and influences the way people
verbs and prepositions (e.g., Göksun et al., 2009; Lakusta understand the language that they hear (for detailed discussions
et al., 2007; McDonough et al., 2003; Pruden, 2006; Pulverman see Gleitman & Papafragou, 2005; Munnich et al., 2001).
et al., 2008). Moreover, and despite the fact that more research
needs to be done, the research suggests that sensitivity to these
constructs is universal in two senses: (a) irrespective of the
Future Questions
language environment in which infants are raised, they detect We are beginning to discover the nonlinguistic constructs
non-linguistic components of events, and (b) infants attend to necessary for the learning of relational terms like verbs and
fine-grained distinctions in events even when these are not prepositions. This article offers a multidisciplinary approach
codified in their native language (Goksun et al., 2008; Hespos to the semantic foundations for language by investigating
& Spelke, 2004). evidence from linguistics, event perception, and language
The second conclusion is that not all conceptual precursors development across four categories of events. For our argu-
emerge at the same time. Infants seem to be able to detect or ments to go through, research on conceptual precursors must
categorize relations of containment before support relations, be broadened to include other categories (e.g., force dynamics,
path before manner, goal before source, and figure before causation, or distance). We suggested that good candidate
ground (e.g., Casasola & Cohen, 2002; Göksun et al., 2008; semantic constructs should be perceptually accessible, univer-
Lakusta et al., 2007; Pruden et al., 2004). This apparent incon- sally seen in the world’s languages, and packaged differently
sistency in the developmental progression may be a function of across languages. Few studies have asked how language influ-
which constructs are expressed more universally than others. ences event perception and whether the trend for infants is from
That is, the more prevalent a distinction is across languages, the universal to language-specific patterns. More cross-linguistic
more likely it is to come early. Rather, the differential studies and studies with bilingual children are necessary to

38
Trading Spaces 39

validate our assertions about trading spaces. Developmental University Infant Lab and University of Delaware Infant Language
patterns across typologically varied languages and ways in Project for their invaluable contributions on various stages of this arti-
which children acquire the biases of their native language will cle. Special thanks to Nora Newcombe, Peter Marshall, and Kim
shed light on the links between language and thought. Curby for helpful comments about this work, Sarah Roseberry for
fruitful discussions, and Katrina Ferrara for detailed edits on earlier
Little is known about the long-term consequences of
drafts of this article.
perceiving and categorizing relations in events. Some of the
research reviewed here is tantalizing for its links to language
Declaration of Conflicting Interests
development. Longitudinal studies examining multiple concep-
tual precursors and their later relations to language develop- The authors declared no conflicts of interest with respect to the
ment, specifically to verb learning, must be conducted. authorship and/or publication of this article.
Preliminary results from ongoing research are promising:
Infants’ ability to categorize foundational components in a non- References
linguistic task of path and manner is correlated with verb learn- Aguiar, A., & Baillargeon, R. (1999). 2.5-month-old infants’ reason-
ing, but not with a nonlinguistic spatial task (Roseberry et al., ing about when objects should and should not be occluded. Cogni-
2009). Finally, the mechanisms underlying relational language tive Psychology, 39, 116–157.
development might provide insights for at least two practical Allen, S., Özyürek, A., Kita, S., Brown, A., Furman, R., Ishizuka, T.,
domains: second language learning and atypical language & Fujii, M. (2007). Language-specific and universal influences in
development. How might educators teach relational terms to children’s syntactic packaging of manner and path: A comparison
students learning a second language when these terms are of English, Japanese, and Turkish. Cognition, 102, 16–48.
notoriously difficult? Perhaps verbs and prepositions would Arterberry, M.E., & Bornstein, M.H. (2002). Infant perceptual and
become more transparent if taught in terms of semantic compo- conceptual categorization: The roles of static and dynamic stimu-
nents (Infiesta, Song, Pulverman, Golinkoff, & Hirsh-Pasek, lus attributes. Cognition, 86, 1–24.
2009). People learning a second language must trade spaces for Bahrick, L.E., & Pickens, J.N. (1995). Infant memory for object
learning relational terms, just as children do. This new perspec- motion across a period of three months: Implications for a four-
tive might have implications for how second language are phase attention function. Journal of Experimental Child Psychol-
taught. ogy, 59, 341–373.
A second domain is atypically developing children’s acqui- Baillargeon, R. (2004). Infants’ physical world. Current Directions in
sition of relational vocabulary. Studies show that children with Psychological Science, 13, 89–94.
autism have delayed language development, particularly in the Baillargeon, R., Needham, A., & DeVos, J. (1992). The development
learning of verbs (Chan, Cheung, Leung, Cheung, & Cheung, of young infants’ intuitions about support. Early Development and
2005). Possibly problems in the learning of relational terms are Parenting, 1, 69–78.
not specifically linguistic in nature but instead stem from diffi- Baillargeon, R., & Wang, S.H. (2002). Event categorization in
culty with finding the precursors for verb meaning in dynamic infancy. Trends in Cognitive Sciences, 6, 85–93.
nonlinguistic events. Our lab is currently investigating this Balaban, M.T., & Waxman, S.R. (1997). Do words facilitate object
question (Parish-Morris, Hirsh-Pasek, & Golinkoff, 2009). categorization in 9-month-old infants? Journal of Experimental
Child Psychology, 64, 3–26.
Baldwin, D.A., Baird, J.A., Saylor, M.M., & Clark, A.M. (2001).
Conclusions Infants parse dynamic action. Child Development, 72, 708–717.
Children’s relational language acquisition has its roots in their Baldwin, D.A., & Markman, E.M. (1989). Establishing word-object
understanding of nonlinguistic spatial and event constructs. relations: A first step. Child Development, 60, 381–398.
These event constructs, represented in all the world’s languages Bertenthal, B.I. (1993). Infants’ perception of biomechanical motions:
although expressed in different ways, are the subject matter of Intrinsic image and knowledge-based constraints. In C. Granrud
the prepositions and verbs that name them. Children appear to (Ed.), Visual perception and cognition in infancy (pp. 175–214).
distinguish between and categorize the components of events in Hillsdale, NJ: Erlbaum.
a somewhat universal way. They then trade spaces based on Bogartz, R.S., Shinskey, J.L., & Schilling, T.H. (2000). Object perma-
how their native language expresses these relations. Thus, just nence in five-and-a-half-month-old infants. Infancy, 1, 403–428.
as language learning narrows children’s phonological space, Booth, A.E., & Waxman, S.R. (2002). Object names and object func-
language exposure promotes semantic reorganization, inclining tions serve as cues to categories for infants. Developmental Psy-
children to focus on those relations that are uniquely packaged chology, 38, 948–957.
by their native language. Booth, A.E., & Waxman, S.R. (2003). Bringing theories of word
learning in line with the evidence. Cognition, 87, 215–218.
Acknowledgments Boroditsky, L. (2001). Does language shape thought? English and Man-
This work was supported by National Institute of Child Health and darin speakers’ conceptions of time. Cognitive Psychology, 43,
Human Development Grant 5R01HD050199 and by National Science 1–22.
Foundation Grants BCS-0642529 to Kathy Hirsh-Pasek and Roberta Boroditsky, L., & Ramscar, M. (2002). The roles of body and mind in
Michnick Golinkoff. We would like to thank everyone at Temple abstract thought. Psychological Science, 13, 185–188.

39
40 Göksun et al.

Bowerman, M., & Choi, S. (1994, November). Linguistic and nonlin- Language acquisition and conceptual development (pp.
guistic determinants of spatial semantic development. Paper pre- 215–256). New York: Cambridge University Press.
sented at the Boston University Conference on Language Gentner, D., & Bowerman, M. (2009). Why some spatial semantic
Development, Boston, MA. categories are harder to learn than others: The typological
Bowerman, M., & Levinson, L. (2001). Language acquisition and prevalence hypothesis. In J. Guo, E. Lieven, S. Ervin-Tripp,
conceptual development. Cambridge, United Kingdom: Cambridge N. Budwig, S. Özcaliskan, & K. Nakamura (Eds.), Crosslinguistic
University Press. approaches to the psychology of language: Research in the
Casasola, M. (2005a). Can language do the driving? The effect of lin- tradition of Dan Isaac Slobin (pp. 465–480). New York: Erlbaum.
guistic input on infants’ categorization of support spatial relations. Gleitman, L., & Papafragou, A. (2005). Language and thought. In
Developmental Psychology, 41, 183–192. R. Morrison & K. Holyoak (Eds.), Cambridge handbook of think-
Casasola, M. (2005b). When less is more: How infants learn to form ing and reasoning (pp. 633–661). Cambridge, United Kingdom:
an abstract categorical representation of support. Child Develop- Cambridge University Press.
ment, 76, 279–290. Göksun, T., Hirsh-Pasek, K., & Golinkoff, R. M. (2008, July). Figure
Casasola, M., & Cohen, L.B. (2002). Infant categorization of contain- and ground: Conceptual primitives for processing events. Paper
ment, support, and tight-fit spatial relationships. Developmental presented in T. Göksun & S. Pruden (Chairs), Foundations for
Science, 5, 247–264. learning relational terms: What is in an event? Symposium at the
Casasola, M., Cohen, L.B., & Chiarello, E. (2003). Six-month-old 11th International Congress for the Study of Child Language, Edin-
infants’ categorization of containment spatial relations. Child burgh, Scotland.
Development, 74, 679–693. Göksun, T., Hirsh-Pasek, K, & Golinkoff, R.M. (2009). Processing fig-
Cashon, C.H., & Cohen, L.B. (2000). Eight-month-old infants’ per- ures and grounds in dynamic and static events. In J. Chandlee,
ception of possible and impossible events. Infancy, 1, 429–446. M. Franchini, S. Lord, & G. Rheiner (Eds.), Proceedings of the
Chan, A.S., Cheung, J., Leung, W.M., Cheung, R., & Cheung, M. 33rd annual Boston University Conference on Language Develop-
(2005). Verbal expression and comprehension deficits in young ment (pp. 199–210). Somerville, MA: Cascadilla Press.
children with Autism. Focus on Autism & Other Developmental Golinkoff, R.M., Chung, H.L., Hirsh-Pasek, K., Liu, J., Bertenthal, B.I.,
Disabilities, 20, 117–124. Brand, R., et al. (2002). Young children can extend motion verb
Choi, S. (2006). Influence of language-specific input on spatial cogni- labels to point-light displays. Developmental Psychology, 38,
tion: Categories of containment. First Language, 26, 207–232. 604–614.
Choi, S., & Bowerman, M. (1991). Learning to express motion events Golinkoff, R.M., & Hirsh-Pasek, K. (2008). How toddlers begin to
in English and Korean: The influence of language-specific lexica- learn verbs. Trends in Cognitive Science, 12, 397–403.
lization patterns. Cognition, 41, 83–121. Hespos, S.J., & Baillargeon, R. (2001a). Infants’ knowledge about
Clark, E.V. (2003). First language acquisition. New York: Cambridge occlusion and containment events: A surprising discrepancy. Psy-
University Press. chological Science, 12, 141–147.
Clark, E.V. (2004). How language acquisitions builds on cognitive Hespos, S.J., & Baillargeon, R. (2001b). Reasoning about containment
development. Trends in Cognitive Sciences, 8, 472–478. events in very young infants. Cognition, 78, 207–245.
Csibra, G., & Gergely, G. (1998). The teleological origins of mentalis- Hespos, S.J., & Baillargeon, R. (2008). Young infants’ actions reveal
tic action explanations: A developmental hypothesis. Developmen- their developing knowledge of support variables: Converging evi-
tal Science, 1, 255–259. dence for violation-of-expectation findings. Cognition, 107,
Csibra, G., Gergely, G., Biro, S., Koos, O., & Brockbank, M. (1999). 304–316.
Goal attribution without agency cues: The perception of ‘‘pure rea- Hespos, S.J., & Piccin, T. (2009). To generalize or not to generalize:
son’’ in infancy. Cognition, 72, 237–267. Spatial categories are influenced by physical attributes and lan-
Eimas, P.D., Miller, J.L., & Jusczyk, P.W. (1987). On infant speech guage. Developmental Science, 12, 88–95.
perception and the acquisition of language. In S. Harnard (Ed.), Cate- Hespos, S.J., & Spelke, E.S. (2004). Conceptual precursors to lan-
gorical perception: The groundwork of cognition (pp. 161–198). guage. Nature, 430, 453–456.
Cambridge, United Kingdom: Cambridge University Press. Hespos, S.J., & Spelke, E.S. (2007). Precursors to spatial language:
Fulkerson, A.L., & Haaf, R.A. (2003). The influence of labels, non- The case of containment. In M. Aurnague, M. Hickman, &
labeling sounds, and source of auditory input on 9- and 15- L. Vieu (Eds.), The categorization of spatial entities in language
month-olds’ object categorization. Infancy, 4, 349–369. and cognition (pp. 233–245). Amsterdam: John Benjamins.
Galles-Sebastian, N. (2006). Native language sensitivities: Evolution Infiesta, C., Song, L., Pulverman, R., Golinkoff, R.M., & Hirsh-
in the first year of life. Trends in Cognitive Science, 10, 239–240. Pasek, K. (2009). Does the owl fly out of the tree or does the owl
Gentner, D. (1982). Why nouns are learned before verbs: Linguistic exit the tree flying? How second language learners cope with
relativity versus natural partitioning. In S. Kuczaj (Ed.), Language encoding events. Manuscript submitted for publication.
development: Language, thought, and culture (Vol. 2, pp. Jackendoff, R. (1983). Semantics and cognition: Current studies in
301–334). Hillsdale, NJ: Erlbaum. linguistics series, No. 8. Cambridge, MA: MIT Press.
Gentner, D., & Boroditsky, L. (2001). Individuation, relativity, and Johanson, M., Selimis, S., & Papafragou, A. (2008, October/
early word learning. In M. Bowerman & S.C. Levinson (Eds.), November). Cross-linguistic biases in the semantics and the

40
Trading Spaces 41

acquisition of spatial language. Paper presented at the 33rd Boston Munnich, E., Landau, B., & Dosher, B. (2001). Spatial language and
University Conference on Language Development, Boston, MA. spatial representation: A cross-linguistic comparison. Cognition,
Kuhl, P.K. (2004). Early language acquisition: Cracking the speech 81, 171–208.
code. Nature Reviews Neuroscience, 5, 831–843. Norbury, H.M., Waxman, S.R., & Song, H. (2008). Tight and loose are
Kuhl, P.K., Andruski, J.E., Chistovich, I.A., Chistovich, L.A., not created equal: An asymmetry underlying the representation of
Kozhevnikova, E.V., & Ryskina, V.L. (1997). Cross-language fit in English- and Korean-speakers. Cognition, 109, 316–325.
analysis of phonetic units in language addressed to infants. Sci- Oakes, L.M. (1994). The development of infants’ use of continuity
ence, 277, 684–686. cues in their perception of causality. Developmental Psychology,
Kuhl, P.K., Conboy, B.T., Padden, D., Nelson, T., & Pruitt, J. (2005). 30, 869–879.
Early speech perception and later language development: Implica- Özçaliskan, S. & Slobin, D.I. (1999). Learning ‘‘how to search for the
tions for the ‘‘critical period.’’ Language Learning and Develop- frog’’: Expression of manner of motion in English, Spanish, and
ment, 1, 237–264. Turkish. In A. Greenhill, H. Littlefield, & C. Tano (Eds.), Proceed-
Lakoff, G. (1987). Women, fire, and dangerous things: What cate- ings of the 23rd annual Boston University Conference on Language
gories reveal about the mind. Chicago: University of Chicago Development (pp. 541–552). Somerville, MA: Cascadilla Press.
Press. Papafragou, A., Hulbert, J., & Trueswell, J. (2008). Does language
Lakusta, L., & Carey, S. (2008, March). Infants’ categorization of guide event perception? Evidence from eye movements. Cogni-
sources and goals in motion events. Paper presented in tion, 108, 155–184.
S. Pruden, & T. Göksun (Chairs), Conceptual primitives for Papafragou, A., Massey, C., & Gleitman, L. (2006). When English
processing events and learning relational terms. Symposium proposes what Greek presupposes: The cross-linguistic encoding
at the 16th International Conference on Infant Studies, of motion events. Cognition, 98, B75–87.
Vancouver, Canada. Parish-Morris, J., Hirsh-Pasek, K., & Golinkoff, R. (2009). Spatial
Lakusta, L., & Landau, B. (2005). Starting at the end: The importance cognition, social understanding, and relational language in chil-
of goals in spatial language. Cognition, 96, 1–33. dren with autism. Manuscript in preparation.
Lakusta, L., Wagner, L., O’Hearn, K., & Landau, B. (2007). Concep- Pruden, S.M. (2006). Finding the action: Factor that aid infants’
tual foundations of spatial language: evidence for a goal bias in abstraction of path and manner. Unpublished dissertation, Temple
infants. Language Learning and Development, 3, 179–197. University, Ambler, PA.
Lakusta, L., Yoshida, H., Landau, B., & Smith, L. (2006, June). Cross- Pruden, S.M., & Hirsh-Pasek, K. (2006). Foundations of verb learn-
linguistic evidence for a goal/source asymmetry: The case of ing: Labels promote action category formation. In D. Bamman,
Japanese. Poster presented at the International Conference on T. Magnitskaia, & C. Zaller (Eds.), Proceedings of the 30th annual
Infant Studies, Kyoto, Japan. Boston University Conference on Language Development (pp.
Langacker, R.W. (1987). Foundations of cognitive grammar. 476–488). Somerville, MA: Cascadilla Press.
Stanford, CA: Stanford University Press. Pruden, S.M., Hirsh-Pasek, K., Maguire, M.J., & Meyer, M.A. (2004).
Leslie, A.M. (1982). The perception of causality in infants. Percep- Foundations of verb learning: Infants form categories of path
tion, 11, 173–186. and manner in motion events. In A. Brugos, L. Micciulla, &
Li, P., Abarbanell, L., Gleitman, L., & Papafragou, A. (2009). Spatial C.E. Smith (Eds.), Proceedings of the 28th annual Boston Univer-
reasoning in Tenejapan Mayans. Manuscript submitted for sity Conference on Language Development (pp. 461–472).
publication. Somerville, MA: Cascadilla Press.
Li, P., & Gleitman, P. (2002). Turning the tables: Language and spa- Pulverman, R., Chen, J., Chan, C., Tardif, T., & Meng, X. (2007,
tial reasoning. Cognition, 83, 265–294. March). Cross-cultural comparisons of attention to manner and
Maguire, M.J., Hirsh-Pasek, K., Golinkoff, R.M., Haryu, E., Imai, M., path: Insights from Chinese infants. Poster presented at the meet-
Vengas, S., et al. (in press). A developmental shift from similar to ing of the Society for Research on Child Development, Boston,
language-specific strategies in verb acquisition: A comparison of MA.
English, Spanish and Japanese. Cognition. Pulverman, R., & Golinkoff, R.M. (2004). Seven-month-olds’ atten-
Mandler, J.M. (1992). How to build a baby: II. Conceptual primitives. tion to potential verb referents in nonlinguistic events. In
Psychological Review, 99, 587–604. A. Brugos, L. Micciulla, & C.E. Smith (Eds.), Proceedings of the
Mandler, J.M. (2004). The foundations of mind: Origins of conceptual 28th annual Boston University Conference on Language Develop-
thought. New York: Oxford University Press. ment (pp. 473–480). Somerville, MA: Cascadilla Press.
McDonough, L., Choi, S., & Mandler, J.M. (2003). Understanding Pulverman, R., Golinkoff, R.M., Hirsh-Pasek, K., & Brandone, A.C.
spatial relations: Flexible infants, lexical adults. Cognitive Psy- (2009). Novel words guide infants’ attention to words. Manuscript
chology, 46, 229–259. submitted for publication.
Muehleisen, V., & Imai, M. (1997). Transitivity and the incorporation Pulverman, R., Golinkoff, R.M., Hirsh-Pasek, K., & Sootsman-
of ground information in Japanese path verbs. In K. Lee, Buresh, J. (2008). Infants discriminate paths and manners in non-
E. Sweetwer, & M. Verspoor (Eds.), Lexical and syntactic con- linguistic dynamic events. Cognition, 108, 825–830.
structions and the construction of meaning (pp. 329–346). Pulverman, R., Hirsh-Pasek, K., Golinkoff, R. M., Pruden, S., &
Amsterdam: John Benjamins. Salkind, S. (2006). Conceptual foundations for verb learning:

41
42 Göksun et al.

Celebrating the event. In K. Hirsh-Pasek & R.M. Golinkoff (Eds.), Talmy, L. (2000). Toward a cognitive semantics: Vol. 1. Concept struc-
Action meets word: How children learn verbs (pp. 134–159). New turing systems. Cambridge, MA: MIT Press 1–565.
York: Oxford University Press. Tees, R.C., & Werker, J.F. (1984). Perceptual flexibility: Maintenance
Regier, T., & Zheng, Z. (2007). Attention to endpoints: A cross- or recovery of the ability to discriminate non-native speech sounds.
linguistic constraint on spatial meaning. Cognitive Science, 31, Canadian Journal of Psychology, 38, 579–590.
705–719. Tomasello, M. (1995). Pragmatic contexts for early verb learning. In
Roseberry, S., Göksun, T., Hirsh-Pasek, K., Newcombe, N., Golink- M. Tomasello & W.E. Merriman (Eds.), Beyond the names for
off, R.M., Novack, M., & Brayfield, R. (2009, April). Individual things: Young children’s acquisition of verbs (pp. 115–146).
differences in early event perception predict later verb learning. Hillsdale, NJ: Erlbaum.
Poster presented at the meeting of the Society for Research on Tsao, F.M., Liu, H.M., & Kuhl, P.K. (2004). Speech perception in
Child Development, Denver, CO. infancy predicts language development in the second year of life:
Sharon, T., & Wynn, K. (1998). Individuation of actions from contin- A longitudinal study. Child Development, 75, 1067–1084.
uous motion. Psychological Science, 9, 357–362. Waxman, S.R. (1999). Specifying the scope of 13-month-olds’ expec-
Shipley T.F., & Zacks, J.M. (Eds.), Understanding events: From per- tations for novel words. Cognition, 70, B35–B50.
ception to action. New York: Oxford University Press. Waxman, S.R., & Booth, A.E. (2003). The origins and evolution of
Slobin, D.I. (1996). From ‘‘thought and language’’ to ‘‘thinking to links between word learning and conceptual organization: New evi-
speaking.’’ In J.J. Gumperz & S.C. Levinson (Eds.), Rethinking dence from 11-month-olds. Developmental Science, 6, 130–137.
linguistic relativity (pp. 70–96). Cambridge, United Kingdom: Werker, J.F., & Lalonde, C.E. (1988). Cross-language speech percep-
Cambridge University Press. tion: Initial capabilities and developmental change. Developmental
Slobin, D.I. (2001). Form-function relations: How do children find out Psychology, 24, 672–683.
what they are? In M. Bowerman & S.C. Levinson (Eds.), Language Werker, J.F., & Tees, R.C. (1984). Cross-language speech perception:
acquisition and conceptual development (pp. 406–449). Evidence for perceptual reorganization during the first year of life.
Cambridge, United Kingdom: Cambridge University Press. Infant Behavior and Development, 7, 49–63.
Song, L., Golinkoff, R.M., Seston, R., Ma, W., Shallcross, W., & Whorf, B.L. (1956). Language, thought, and reality. Cambridge, MA:
Hirsh-Pasek, K. (2006, November). Action stations: Verb learning MIT Press.
rests on constructing categories of action. Poster presented at the Woodward, A.L. (1998). Infants selectively encode the goal object of
31st Boston University Conference on Language Development, an actor’s reach. Cognition, 69, 1–34.
Boston, MA. Woodward, A. (1999). Infants’ ability to distinguish between purpose-
Spelke, E.S., Born, W.S., & Chu, F. (1983). Perception of moving, sound- ful and non-purposeful behaviors. Infant Behavior & Development,
ing objects by four-month-old infants. Perception, 12, 719–732. 22, 145–160.
Spelke, E.S. & Hespos, S.J. (2002). Conceptual development in Wynn, K. (1996). Infants’ individuation and enumeration of actions.
infancy: The case of containment. In N. Stein, P. Bauer, & Psychological Science, 7, 164–169.
M. Rabinowitch (Eds.), A festschrift for Jean Mandler (pp. Zacks, J.M., & Tversky, B. (2001). Event structure in perception and
223–246). Hillsdale, NJ: Erlbaum. conception. Psychological Bulletin, 127, 3–21.
Talmy, L. (1985). Lexicalization patterns: Semantic structure in lexi- Zheng, M.Y., & Goldin-Meadow, S. (2002). Thought before language:
cal forms. In T. Shopen (Ed.), Language typology and syntactic How deaf and hearing children express motion events across cul-
description (pp. 57–149). New York: Cambridge University Press. tures. Cognition, 85, 145–175.

42