Emerging Trends in The Social and Behavioral Sciences

Language, Perspective, and Memory

Item

Title

Language, Perspective, and Memory

Author

Ryskin, Rachel A.

Yoon, Si On

Brown‐Schmidt, Sarah

Research Area

Cognition and Emotions

Topic

Language

Abstract

The ability to take the perspective of another person is ubiquitous in many everyday cognitive activities. In particular, it allows people to communicate efficiently with conversational partners. Speakers tailor what they say based on the listener's knowledge and, likewise, listeners use what they know about the speaker to better understand what the speaker means. In this essay, we review foundational research on the role of perspective‐taking in the domain of language processing and describe new lines of work that are beginning to explore the memory processes that support the efficient use of perspectives in conversation. We then discuss key avenues for future research, such as investigating whether the type of perspective‐taking involved in creating memory reminders draws on the same underlying cognitive processes as in the domain of language processing. Exploring this interface between language, perspective‐taking, and memory will require interdisciplinary crosstalk and integration of methodologies across the domains of memory and language research.

Related Essays

Theory of Mind and Behavior (Psychology), Amanda C. Brandone

Delusions (Psychology), Max Coltheart

Misinformation and How to Correct It (Psychology), John Cook et al.

Insight (Psychology), Brian Erickson and John Kounios

Cognitive Processes Involved in Stereotyping (Psychology), Susan T. Fiske and Cydney H. Dupree

Language and Thought (Psychology), Susan Goldin‐Meadow

Concepts and Semantic Memory (Psychology), Barbara C. Malt

Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg

Gestural Communication in Nonhuman Species (Anthropology), Simone Pika

Attention and Perception (Psychology), Ronald A. Rensink

Vocal Communication in Primates (Anthropology), Katie E. Slocombe

How Form Constrains Function in the Human Brain (Psychology), Timothy D. Verstynen

Speech Perception (Psychology), Athena Vouloumanos

Theory of Mind (Psychology), Henry Wellman

Identifier

etrds0200

extracted text

Language, Perspective, and Memory
RACHEL A. RYSKIN, SI ON YOON, and SARAH BROWN-SCHMIDT

Abstract
The ability to take the perspective of another person is ubiquitous in many everyday
cognitive activities. In particular, it allows people to communicate efficiently with
conversational partners. Speakers tailor what they say based on the listener’s knowledge and, likewise, listeners use what they know about the speaker to better understand what the speaker means. In this essay, we review foundational research on the
role of perspective-taking in the domain of language processing and describe new
lines of work that are beginning to explore the memory processes that support the
efficient use of perspectives in conversation. We then discuss key avenues for future
research, such as investigating whether the type of perspective-taking involved in
creating memory reminders draws on the same underlying cognitive processes as
in the domain of language processing. Exploring this interface between language,
perspective-taking, and memory will require interdisciplinary crosstalk and integration of methodologies across the domains of memory and language research.

INTRODUCTION
Perspective-taking is the ability to mentally represent the beliefs and knowledge of another person, which may or may not differ from one’s own. More
generally, the capacity to appreciate alternative perspectives is a key contributor to a variety of human cognitive activities—from being able to read a
map (Figure 1) to interacting with another person. Consider that when any
two individuals come together, they have a certain amount of knowledge
about the world that is shared between them. For two Americans sitting in
a ball-park watching a game, this joint knowledge or common ground would
include popular cultural references such as the name of the President of the
United States, as well as immediately available information in the context
such as information about the last play, or the score in the game. These individuals would also have a certain amount of information that is private to
each individual, and not shared with the other person, for example, what
each one of them ate for breakfast. A central goal of many interactions is
sharing some of this private knowledge, sometimes called privileged ground, in
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

School

Home

A

B
Toy store

Figure 1 A and B are both reading the same map from opposite perspectives.
For A, the school is on the left. For B, the school is on the right. If B is directing A
about how to walk from home to the school, B will have to take the opposite
perspective (or vice versa); otherwise, A might end up at the toy store.

order to introduce new information and keep the conversation interesting
(after all, a conversation is bound to be painfully boring if it only involved
repeating to one another what is already jointly known).
Keeping track of what others do and do not know is an ability that is
observed even in very young children. For example, when young children
see an actor place a toy inside a container and leave the room, they expect
the actor, on her return, to search for the toy in that same container even
if the child knows that the toy was removed from this container. Thus, the
child is able to distinguish their privileged knowledge that the toy is no
longer in the container, from the actor’s false belief that the toy is in the
container. This ability to distinguish self-knowledge from other-knowledge
allows the child to generate the prediction that the actor will search for her
toy in the original location. The ability to represent the fact that the actor
has a false belief about the location of the toy—a belief that differs from the
child’s own—is present by the second year of life, if not earlier (Baillargeon,
Scott, & He, 2010).
Perspective-taking is likewise a key contributor to adult interpersonal communication; effective speakers tailor how they speak to their audience, and
effective listeners adjust their understanding of what is said based on what
they know about their interlocutor (Clark, 1996). For example, when describing New York City landmarks, New Yorkers will use proper names (e.g.,
Rockefeller Center) when speaking to other New Yorkers and descriptive

Language, Perspective, and Memory

3

phrases (e.g., the building with the flags in front of it) when speaking to a
person who has never been to New York (Isaacs & Clark, 1987). Similarly, an
effective listener might interpret a word such as tweet with different meanings, depending on their beliefs about what the speaker knows. If a college
student were listening to a young person, tweet might bring to mind the social
media site, “Twitter”; if they were talking to an elderly person, the same word
might bring to mind a bird.
This ability to adjust our expectations about the actions of others and to tailor how we use language with other individuals critically depends on representations of the perspective of other individuals. In order to appreciate that
another person has a different perspective than one’s own, for example, individuals must store in memory a representation of their interlocutor’s knowledge state that is distinct from their own knowledge state. They must also
be able to access those memory representations quickly, when cued appropriately. The nature of these memory structures and the ways in which they
are accessed during language production and comprehension are areas of
active research. The products of this research have implications for a variety
of cognitive domains.
FOUNDATIONAL RESEARCH
Representations of the perspective of others play a central role in face-to-face
communicative settings. In conversation, the common ground between the
interlocutors is thought to be the fundamental backdrop against which communication takes place (Clark, 1996). As a result, many of the advances in
our understanding of how perspective is represented and used come from
psycholinguistic studies of how language is processed in face-to-face conversation. This research shows that both speakers and listeners alike tailor
language processes based on the knowledge that is jointly held with their
conversational partner.
COMMON GROUND IN LANGUAGE PRODUCTION
It is well known that when we refer to something in the world, we must
distinguish what we intend to describe from other potential referents (Olson,
1970; Osgood, 1971). When one orders a doughnut at a favorite doughnut
shop, it simply will not suffice to say “One doughnut, please.” Instead, the
speaker must specify her referential expression “one doughnut” to pick
out the one she wants, from many that she does not want, as in “One
glazed chocolate cake doughnut, please.” Thus, speakers design their referring
expressions with respect to the physical and cognitive environment they find
themselves in. Consistent with the logic that common ground is central to

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

language use (Clark, 1996), speakers also change their utterances depending
on their beliefs about what their conversational partner knows and does
not know. For example, imagine a situation in which a speaker will ask a
listener to hand her one of several items in a display, such as the smaller of
two triangles in Figure 2. In a situation where both the speaker and listener
see the two triangles, speakers typically specify the size of the triangle, as
in “Please hand me the small triangle”(Figure 2a). The adjective “small” plays
a critical role in this linguistic exchange, because it specifies which of the
two triangles the speaker is referring to. However, if the larger triangle
is not visible to their partner (Figure 2b), speakers often will refer to it
simply as “the triangle” because the contrasting adjective, “small,” becomes
superfluous (Nadig & Sedivy, 2002; Wardlow Lane, Groisman, & Ferreira,
2006) and potentially confusing. From the listener’s perspective, there is
only one triangle. Thus, speakers use their knowledge about the listener’s
perspective in the situation, in order to constrain what they say.
While the shared physical environment is one component of common
ground, another mechanism by which common ground is formed is through
communication itself. A central finding in studies of conversational language
is that, as we converse, the ways in which we refer to various topics of discussion become more definite and succinct over the course of a conversation
(Wilkes-Gibbs & Clark, 1992). For instance, when describing an ambiguous
black-and-white image such as the leftmost object in Figure 3, on a first
attempt, a speaker might say, “it resembles someone who looks like they’re trying
to climb stairs. There’s two feet, one is way above the other.” When redescribing
the same figure later in the conversation, the speaker might now refer to
it more concretely and efficiently, as in “the stair climber.” Memory for the
previous mention of that ambiguous figure, and the knowledge that this
memory is shared with one’s conversational partner, allows the speaker
to refine her phrasing. This phenomenon is observable even on a global
level. Consider that in the early days of the Internet, commercials enticed
potential customers to view their product on the “world wide web,” whereas
nowadays, we simply refer to the “web.” This shortening of the expression

Listener

(a)

Speaker

Listener

(b)

Speaker

Figure 2 (a) Both triangles are visible to the listener. (b) A barrier blocks the larger
triangle, making it hidden from the listener’s perspective.

Language, Perspective, and Memory

Figure 3

5

Example tangrams.

represents the growth in global common knowledge about the Internet, and,
as such, indicates that, as a global community, we have formed common
ground for this concept.
Key evidence in support of the argument that these phenomena represent
common ground between people (and not simply private knowledge, e.g.,
about what the web is) comes from findings that speakers no longer use these
shortened terms such as stair climber or web when speaking with a person who
is unknowledgeable about the term. In such circumstances, speakers revert
to longer, more descriptive phrases (Horton & Gerrig, 2002; Wilkes-Gibbs &
Clark, 1992). This ability to tailor referential expressions to the knowledge of
specific addressees is remarkably flexible and powerful. Speakers are able to
keep track of a large number of distinct and intermixed items, some of which
are shared with one partner, and some of which are shared with a different
partner (Horton & Gerrig, 2005). Speakers are also able to switch back and
forth between conversational partners with whom they share distinct knowledge, appropriately adjusting their expressions to be more descriptive when
the current addressee is “naïve” with respect to that given item, and using
established naming conventions when the current addressee is familiar with
the convention (Horton & Gerrig, 2002; Yoon & Brown-Schmidt, 2014).
COMMON GROUND IN LANGUAGE COMPREHENSION
In much the same way as speakers tailor how they speak depending on
the knowledge of their addressee, listeners, too, engage in complementary,
partner-specific adjustment during conversation. Once two people in a
conversation have converged on a particular expression (e.g., “the shiny
cylinder”) to refer to a given referent, listeners are surprised to hear the
same speaker switch to an equally plausible but novel expression (e.g., “the
silver pipe”), suggesting that the novel expression provides a less effective
cue to the representation of that referent stored in memory (Metzing &
Brennan, 2003). However, if a new person joins the conversation, these
novel expressions are no longer surprising, and are processed more quickly.

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

The fact that the penalty for novel expressions does not extend to speakers
who do not share common ground for the original term, shows that this
is not simply an egocentric process. Instead, listeners take into account
what knowledge their partner does and does not have when generating an
interpretation of what they say.
Listeners similarly use what they know about a speaker’s visual perspective
within the physical context when interpreting language (Hanna, Tanenhaus,
& Trueswell, 2003; Heller, Grodner, & Tanenhaus, 2008; Nadig & Sedivy,
2002). For instance, if a listener is looking at a display that includes two red
triangles, only one of which is visible to the speaker (Figure 4), when the
listener hears an instruction such as, “Click on the red triangle,” the listener
will look primarily at the red triangle in common ground. The listener is
able to rule out the triangle in privileged ground because she knows that,
from the speaker’s perspective, only one triangle is visible. Complementary
findings include evidence that if a speaker asks you to hand her something,
e.g., “Hand me the cake mix … ,” listeners interpret the speaker as requesting
the cake mix that is out of her own reach, and not a cake mix that she
could have reached from her position; after all, she would not need to ask
for it if it were within reach (Hanna & Tanenhaus, 2004). Further, when
interpreting informational questions, for example, “What’s above the … ,”
listeners readily interpret them as asking about something that the speaker
does not already know—otherwise, there is no need to ask a question

Figure 4 Listener’s view while listening to the instruction, “Click on the red
triangle.” The gray background indicates an object that is visible to the listener, but
not to the speaker.

Language, Perspective, and Memory

7

in the first place (Brown-Schmidt, 2012; Brown-Schmidt, Gunlogson, &
Tanenhaus, 2008). Taken together, this body of research shows that listeners
access representations of the speaker’s perspective, and incorporate this
information into the unfolding interpretation of the speaker’s utterances.
One way of conceptualizing this sensitivity to partner-specific common
ground is as a type of contextual integration, in which information, such as a
jointly established description of a referent (e.g., the shiny cylinder), is bound
to the person with whom the information was experienced. Framed in this
way, understanding how this binding of partner identity with information
from the conversation is encoded, maintained, and accessed in the service
of language processing becomes a central question. Research in the memory
tradition distinguishes concepts such as item memory, including memory
for a word on a list, from the source of that information, including the
room the words were studied in (see Johnson, Hashtroudi, & Lindsay,
1993, for a review). While source in the memory literature is conceptualized
broadly as the conditions in which information is encoded, in the domain of
conversational language, perhaps the most important source is the identity
and perspective of the person with whom you are conversing. Consistent
with this idea that interlocutors are particularly good contextual cues comes
from a meta-analysis of environmental context effects in memory research
(e.g., studies in which the locations of study and test are either the same
or different), which reported a large effect of whether the Experimenter
stayed the same from study to test (Smith & Vela, 2001). In what follows,
we discuss new research that furthers our understanding of the role of
context in a discourse and explores the neural underpinnings of discourse
representations. We then describe potential domains of inquiry that could
be pursued to fruitfully expand our understanding of the link between
memory and language as it pertains to perspective-taking.
CUTTING-EDGE RESEARCH
The literature on language processing in conversation has clearly established
that information in the local context guides language use, including visually
available information (Olson, 1970), as well as information about what
a partner does and does not see or know (Hanna et al., 2003; Nadig &
Sedivy, 2002). While it is clear that conversational partners encode and use
information about content that is discussed, such as the way particular
referents are described (Wilkes-Gibbs & Clark, 1992), less well established
is the extent to which contextual information, including information about
source, is encoded and maintained in conversation. Indeed, source memory
is often more fragile than item memory (e.g., Ferguson, Hashtroudi, &
Johnson, 1992).

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

An important question then is whether memory for source can be cued
effectively at later time points in a conversation, such that future referring is
altered as a result of previous contextual information, or whether the change
of temporal and cognitive context is enough to render the source information inaccessible. Moreover, what are the memory systems that support the
encoding and maintenance of item and source information to be accessed in
the discourse?
PAST CONTEXTS AND FUTURE LANGUAGE USE
According to the context-memory view of discourse (Yoon & Brown-Schmidt,
2013), speakers maintain memory for how they had described previous referents in the conversation. These memory representations guide not only how
speakers re-refer to these referents (Brennan & Clark, 1996; Wilkes-Gibbs &
Clark, 1992) but also how they describe other, previously unnamed referents
in the future (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013). Key evidence for persistent effects of past contexts on future referring comes from
experiments that examine how the way in which speakers refer to a given
referent in one context, shapes how the same item is referred to in a new context. For example, if a pair of conversational partners are presented with a
display that contains several different fish pictures (Figure 5a), and are asked
to refer to each picture, partners will establish brief but descriptive labels for
each fish (e.g., “the curved round fish”) that distinguish among the category
exemplars in the scene (i.e., the other two fish).
Evidence of persistent context effects comes from findings that even when
the context changes and now includes only a single category exemplar
(Figure 5b), speakers continue to use previously established expressions
(e.g., “the curved round fish”) even though they are overly specific given the
immediate context. In other words, while the abbreviated expression the fish

(a)

(b)

(c)

Figure 5 (a) Speakers say “the curved round fish” to disambiguate the target
(circled) fish from the other two fish. (b) Speakers say “the curved round fish” to
disambiguate the target fish from the contrasting fish seen previously (lexical
entrainment). (c) Speakers say “the glittery fish” to disambiguate the novel target
fish from previously discussed fish (lexical differentiation effect).

Language, Perspective, and Memory

9

would have uniquely identified the referent, speakers persist in using the
longer, more descriptive term that had been established previously (Brennan
& Clark, 1996). According to the context-memory view, the source of this
effect is that the speaker’s representation of the relevant context includes
both the single fish in the immediate visual context as well as multiple other
contrasting fish from past contexts.
Distinguishing this context-memory view from a simpler account on which
speakers had simply encoded the term the curved round fish as the name of
the referent (rather than encoding both the term and the items in the past
context), is supported by evidence that previous referents shape the descriptions of future referents. One demonstration of this is the phenomenon of
lexical differentiation (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), in
which speakers differentiate new discourse referents from similar referents
discussed in past contexts. For example, imagine a different situation in
which conversational partners first refer to the fish in Figure 5b as the fish. In
a subsequent context in which a different, novel fish is present (Figure 5c),
speakers are more likely to describe it with a modified noun phrase (e.g.,
glittery fish) compared to a situation where they had not previously described
other fish (Van der Wege, 2009). This lexical differentiation effect shows
that speakers consider how distinct referents from the discourse history
were previously described when designing new expressions. One intriguing
finding is that while Yoon and Brown-Schmidt (2013) observed lexical
differentiation in language production, there was no evidence that listeners
expected speakers to differentiate current from past referents. Instead,
listeners interpreted established terms such as fish quickly, regardless of
whether the term had previously been used to describe a different fish
or not. This suggests that the scope of the relevant discourse context for
speakers and listeners may differ, resulting in subtle asymmetries in the
relationship between reference and context.
MEMORY SYSTEMS IN LANGUAGE USE
The persistence of the historical discourse context in language use raises the
question of how this information is encoded in memory. At least some evidence suggests that multiple memory systems may be involved. In particular, neuropsychological work with individuals who have severe declarative
memory impairment (amnesia) due to bilateral hippocampal damage shows
patterns of deficits and sparing of function consistent with contributions of
multiple memory systems (Duff & Brown-Schmidt, 2012). The idea that the
hippocampus might be involved in processing partner-specific representations of common ground falls naturally from other findings pointing to hippocampal involvement in the processing of object–location relations, even

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

over short time-scales (Hannula, Tranel, & Cohen, 2006; Hannula & Ranganath, 2008). For example, individuals with amnesia are largely successful
at using information about what a conversational partner can and cannot see
in a visual display to constrain interpretation of referring expressions (Rubin,
Brown-Schmidt, Duff, Tranel, & Cohen, 2011). However, when the same individuals were presented with situations in which common ground was established linguistically (and no visual cues to what was shared and what was
not shared were present), they were no longer able to distinguish common
ground from privileged ground after a brief delay. Converging evidence from
studies of the ability to process short narratives reveals significant deficits in
the ability of individuals with amnesia to track discourse referents over the
course of a story, and use information about their relative salience to process pronouns that refer back to these characters (Kurczek, Brown-Schmidt, &
Duff, 2013). Yet, other evidence shows sparing of learning. Specifically, individuals with amnesia successfully develop talker-specific representations of
the accents of different talkers, and use these representations to guide speech
perception (Trude, Duff, & Brown-Schmidt, 2014). Together, these findings
emphasize the need to understand the memory systems that contribute to
and are necessary for the ability to tailor language use to one’s conversational
partner.
KEY ISSUES FOR FUTURE RESEARCH
The literature on language processing in conversation has clearly established
that better understanding the relationship between language and memory
processes is a key area in need of development, not only in terms of experimental findings but also in terms of theory building and the design of implemented models of the underlying processes. In what follows, we describe
several new and promising lines of inquiry that address these questions.
According to one recent proposal, the cognitive processes involved in
the design of a referring expression for another individual are analogous
to those in play when one is constructing a cue that will be used later to
access a specific event in memory (Tullis, 2013; Tullis & Benjamin, 2014).
In a cue-generation task, learners are given a large set of target words and
told to generate one or more cue words for each one. These cues are then
presented to the participants after a delay and they are told to retrieve the
target words they had previously read. Tullis found that learners generate
cues differently when they know that they will be used for future retrieval
compared to when they are simply asked to generate descriptions. They also
make their cues less distinctive and more strongly associated with the target
item when they are generating cues for another person compared to when
they are generating cues for themselves.

Language, Perspective, and Memory

11

The adjustments made during cue generation might reflect underlying processes similar to those occurring during perspective-taking in language use.
For instance, when taking notes in a class, the student’s goal is to jot down
short reminders that will, on later reading, trigger the recollection of the
larger, more complex piece of information that needs to be learned. It can
be challenging to create such reminders. Often they are not successful in eliciting the appropriate recollection (e.g., when the student rereads her notes
and cannot remember what they meant). In generating these cues, the student must essentially take the perspective of her future self and write notes
tailored to her future self’s perspective and knowledge state. Similarly, when
rereading notes, the student must use her knowledge of her past perspective
to constrain the possible interpretations of the cues.
An important question for future work is whether these two seemingly
analogous tasks—perspective-taking during language processing and during cue generation—do, in fact, reflect the same cognitive mechanisms, and,
if so, what the underlying domain-general cognitive functions might be. For
example, some work suggests that individuals who score higher on a test of
working memory are better at designing referring expressions based on the
perspective of their partner (Wardlow, 2013). If similar domain-general mechanisms such as working memory also confer specific benefits in effective cue
generation, this would make the prediction that individuals who are successful in one domain might also be successful in the other. This has important
implications for understanding the architecture of the cognitive mechanisms
underlying language processing and memory. Further, better understanding
how these fundamental skills are interconnected may lead to more strategic
interventions for those who may have deficits in either domain (e.g., older
adults).
Another important line of inquiry will be to identify how a participant’s
role in a conversation, for example, as speaker, listener, or overhearer,,
contributes to the encoding of conversational contextual cues. Recall that
Yoon and Brown-Schmidt (2013) found that while speakers showed sensitivity to the historical discourse context resulting in lexical differentiation
effects, listeners did not. This finding suggests that the representation of
the discourse record might differ for speakers and listeners. If so, critical
questions remain about how the way in which the historical discourse is
encoded in conversation—whether it is from a speaker’s perspective or a
listener’s—guides how this information is used. Conversational partners
are thought to encode contextually rich representations of joint experiences
(Clark & Marshall, 1978), possibly through automatic memory processes that
associate partners with experienced information (Horton & Gerrig, 2005),
and recent findings point to a tight degree of coupling in the representations
of speakers and listeners (Brown-Schmidt & Tanenhaus, 2008; Richardson,

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Dale, & Kirkham, 2007). Thus, understanding the extent to which coordinated representations are normative, and when representations are not
coordinated, would seem to be an important area of inquiry. Related to this
question is that of the breadth of the encoding of the discourse context. For
example, while it is clear that speakers maintain, and use, representations of
previous referents (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), an
open question is whether speakers consider other aspects of the previous
discourse context as well, such as unmentioned properties of previous
referents, or unmentioned objects in the discourse context. On the basis of
the findings in the memory tradition, the degree to which such contextual
information is encoded may depend on whether the encoding and test
conditions emphasize integration of the information with the context, or
instead whether associations among discussed items are more important
(see Eich, 1985; Smith & Vela, 2001).
As we move closer to the goal of developing theories of conversational
memory and language use, such efforts will undoubtedly benefit from
increased collaboration and interaction among researchers in both memoryand language-processing traditions. Research on conversation that combines
the visual world eye-tracking technique (Tanenhaus, Spivey-Knowlton,
Eberhard, & Sedivy, 1995), well established as a key method in language
processing research because it provides valuable information about the time
course of comprehension processes, with methods such as explicit memory
measures and signal detection analyses (Banks, 1970), will likely prove
invaluable in these efforts. In addition, the emerging trend of using Amazon’s Mechanical Turk to collect data about learning and language use over
time (e.g., Fine & Jaeger, 2013) will likely become an increasingly important
resource as it affords collecting large amounts of data from representative
populations. These new approaches, along with a more integrative view of
conversational language research, one that melds language and memory
perspectives (e.g., Stafford & Daly, 1984), will contribute to formulating a
more unified framework of these cognitive processes.

REFERENCES
Baillargeon, R., Scott, R. M., & He, Z. (2010). False-belief understanding in infants.
Trends in Cognitive Sciences, 14, 110–118.
Banks, W. P. (1970). Signal detection theory and human memory. Psychological Bulletin, 74, 81–99.
Brennan, S. E., & Clark, H. H. (1996). Conceptual pacts and lexical choice in conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22,
482–493.

Language, Perspective, and Memory

13

Brown-Schmidt, S. (2012). Beyond common and privileged: Gradient representations
of common ground in real-time language use. Language and Cognitive Processes, 27,
62–89.
Brown-Schmidt, S., Gunlogson, C., & Tanenhaus, M. K. (2008). Addressees distinguish shared from private information when interpreting questions during interactive conversation. Cognition, 107, 1122–1134.
Brown-Schmidt, S., & Tanenhaus, M. K. (2008). Real-time investigation of referential
domains in unscripted conversation: A targeted language game approach. Cognitive Science, 32, 643–684.
Clark, H. H. (1996). Using language. Cambridge, England: Cambridge University
Press.
Clark, H. H., & Marshall, C. R. (1978). Reference diaries. In D. L. Waltz (Ed.), Theoretical issues in natural language processing (Vol. 2, pp. 57–63). New York, NY: Association for Computing Machinery.
Duff, M. C., & Brown-Schmidt, S. (2012). The hippocampus and the flexible use and
processing of language. Frontiers in Cognitive Science, 6, 1–9.
Eich, E. (1985). Context, memory, and integrated item/context imagery. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 11, 764–770.
Ferguson, S. A., Hashtroudi, S., & Johnson, M. K. (1992). Age differences in using
source-relevant cues. Psychology and Aging, 7, 443.
Fine, A. B., & Jaeger, T. F. (2013). Syntactic priming in language comprehension
allows linguistic expectations to converge on the statistics of the input. In M.
Knauff, M. Pauen, N. Sebanz & I. Wachsmuth (Eds.), Proceedings of the 35th Annual
Meeting of the Cognitive Science Society (pp. 2279–2284). Austin, TX: Cognitive Science Society.
Hanna, J. E., & Tanenhaus, M. K. (2004). Pragmatic effects on reference resolution in
a collaborative task: Evidence from eye movements. Cognitive Science, 28, 105–115.
Hanna, J. E., Tanenhaus, M. K., & Trueswell, J. C. (2003). The effects of common
ground and perspective on domains of referential interpretation. Journal of Memory
and Language, 49, 43–61.
Hannula, D. E., & Ranganath, C. (2008). Medial temporal lobe activity predicts successful relational memory binding. The Journal of Neuroscience, 28, 116–124.
Hannula, D. E., Tranel, D., & Cohen, N. J. (2006). The long and the short of it: Relational memory impairments in amnesia, even at short lags. The Journal of Neuroscience, 26, 8352–8359.
Heller, D., Grodner, D., & Tanenhaus, M. K. (2008). The role of perspective in identifying domains of reference. Cognition, 108, 831–836.
Horton, W. S., & Gerrig, R. J. (2002). Speakers’ experiences and audience design:
Knowing when and knowing how to adjust utterances to addressees. Journal of
Memory and Language, 47, 589–606.
Horton, W. S., & Gerrig, R. J. (2005). The impact of memory demands on audience
design during language production. Cognition, 96, 127–142.
Isaacs, E. A., & Clark, H. H. (1987). References in conversation between experts and
novices. Journal of Experimental Psychology: General, 116, 26–37.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3.
Kurczek, J., Brown-Schmidt, S., & Duff, M. (2013). Hippocampal contributions to
language: Evidence of referential processing deficits in amnesia. Journal of Experimental Psychology: General, 142, 1346–1354.
Metzing, C., & Brennan, S. E. (2003). When conceptual pacts are broken: Partnerspecific effects on the comprehension of referring expressions. Journal of Memory
and Language, 49, 201–213.
Nadig, A. S., & Sedivy, J. C. (2002). Evidence of perspective-taking constraints in
children’s on-line reference resolution. Psychological Science, 13, 329–336.
Olson, D. R. (1970). Language and thought: Aspects of a cognitive theory of semantics. Psychological Review, 77, 257–273.
Osgood, C. E. (1971). Exploration in semantic space: A personal diary. Journal of Social
Issues, 27, 5–64.
Richardson, D. C., Dale, R., & Kirkham, N. Z. (2007). The art of conversation is coordination common ground and the coupling of eye movements during dialogue.
Psychological Science, 18, 407–413.
Rubin, R. D., Brown-Schmidt, S., Duff, M. C., Tranel, D., & Cohen, N. J. (2011).
How do I remember that I know you know that I know? Psychological Science, 22,
1574–1582.
Smith, S. M., & Vela, E. (2001). Environmental context-dependent memory: A review
and meta-analysis. Psychonomic Bulletin & Review, 8, 203–220.
Stafford, L., & Daly, J. A. (1984). Conversational memory. Human Communication
Research, 10, 379–402.
Tanenhaus, M. K., Spivey-Knowlton, M. J., Eberhard, K. M., & Sedivy, J. E. (1995).
Integration of visual and linguistic information in spoken language comprehension. Science, 268, 1632–1634.
Trude, A. M., Duff, M., & Brown-Schmidt, S. (2014). Talker-specific learning in amnesia: Insight into mechanisms of adaptive speech perception. Cortex, 54, 117–123.
Tullis, J. G. (2013). Cue generation: How learners flexibly support future retrieval (Unpublished doctoral dissertation). University of Illinois at Urbana-Champaign, Illinois.
Tullis, J. G., & Benjamin, A. S. (in press). Cueing others’ memories. Memory & Cognition.
Van Der Wege, M. M. (2009). Lexical entrainment and lexical differentiation in reference phrase choice. Journal of Memory and Language, 60, 448–463.
Wardlow, L. (2013). Individual differences in speakers’ perspective taking: The roles
of executive control and working memory. Psychonomic Bulletin & Review, 20,
766–772.
Wardlow Lane, L., Groisman, M., & Ferreira, V. S. (2006). Don’t talk about pink
elephants! Speaker’s control over leaking private information during language
production. Psychological Science, 17, 273–277.
Wilkes-Gibbs, D., & Clark, H. H. (1992). Coordinating beliefs in conversation. Journal
of Memory and Language, 31, 183–194.
Yoon, S. O., & Brown-Schmidt, S. (2013). Lexical differentiation in language production and comprehension. Journal of Memory and Language, 69, 397–416.

Language, Perspective, and Memory

15

Yoon, S. O., & Brown-Schmidt, S. (2014). Adjusting conceptual pacts in three-party
conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition,
40, 919–937.

RACHEL A. RYSKIN SHORT BIOGRAPHY
Rachel A. Ryskin is a graduate student in the Department of Psychology at
the University of Illinois at Urbana Champaign, rryskin@gmail.com.
SI ON YOON SHORT BIOGRAPHY
Si On Yoon is a graduate student in the Department of Psychology at the
University of Illinois at Urbana Champaign, sion0912@gmail.com.
SARAH BROWN-SCHMIDT SHORT BIOGRAPHY
Sarah Brown-Schmidt is an Assistant Professor in the Department of
Psychology at the University of Illinois at Urbana Champaign, sarahbrownschmidt@gmail.com. Webpage: sarahbrownschmidt.com/professional.
RELATED ESSAYS
Theory of Mind and Behavior (Psychology), Amanda C. Brandone
Delusions (Psychology), Max Coltheart
Misinformation and How to Correct It (Psychology), John Cook et al.
Insight (Psychology), Brian Erickson and John Kounios
Cognitive Processes Involved in Stereotyping (Psychology), Susan T. Fiske and
Cydney H. Dupree
Language and Thought (Psychology), Susan Goldin-Meadow
Concepts and Semantic Memory (Psychology), Barbara C. Malt
Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Attention and Perception (Psychology), Ronald A. Rensink
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
Speech Perception (Psychology), Athena Vouloumanos
Theory of Mind (Psychology), Henry Wellman



Language, Perspective, and Memory
RACHEL A. RYSKIN, SI ON YOON, and SARAH BROWN-SCHMIDT

Abstract
The ability to take the perspective of another person is ubiquitous in many everyday
cognitive activities. In particular, it allows people to communicate efficiently with
conversational partners. Speakers tailor what they say based on the listener’s knowledge and, likewise, listeners use what they know about the speaker to better understand what the speaker means. In this essay, we review foundational research on the
role of perspective-taking in the domain of language processing and describe new
lines of work that are beginning to explore the memory processes that support the
efficient use of perspectives in conversation. We then discuss key avenues for future
research, such as investigating whether the type of perspective-taking involved in
creating memory reminders draws on the same underlying cognitive processes as
in the domain of language processing. Exploring this interface between language,
perspective-taking, and memory will require interdisciplinary crosstalk and integration of methodologies across the domains of memory and language research.

INTRODUCTION
Perspective-taking is the ability to mentally represent the beliefs and knowledge of another person, which may or may not differ from one’s own. More
generally, the capacity to appreciate alternative perspectives is a key contributor to a variety of human cognitive activities—from being able to read a
map (Figure 1) to interacting with another person. Consider that when any
two individuals come together, they have a certain amount of knowledge
about the world that is shared between them. For two Americans sitting in
a ball-park watching a game, this joint knowledge or common ground would
include popular cultural references such as the name of the President of the
United States, as well as immediately available information in the context
such as information about the last play, or the score in the game. These individuals would also have a certain amount of information that is private to
each individual, and not shared with the other person, for example, what
each one of them ate for breakfast. A central goal of many interactions is
sharing some of this private knowledge, sometimes called privileged ground, in
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

School

Home

A

B
Toy store

Figure 1 A and B are both reading the same map from opposite perspectives.
For A, the school is on the left. For B, the school is on the right. If B is directing A
about how to walk from home to the school, B will have to take the opposite
perspective (or vice versa); otherwise, A might end up at the toy store.

order to introduce new information and keep the conversation interesting
(after all, a conversation is bound to be painfully boring if it only involved
repeating to one another what is already jointly known).
Keeping track of what others do and do not know is an ability that is
observed even in very young children. For example, when young children
see an actor place a toy inside a container and leave the room, they expect
the actor, on her return, to search for the toy in that same container even
if the child knows that the toy was removed from this container. Thus, the
child is able to distinguish their privileged knowledge that the toy is no
longer in the container, from the actor’s false belief that the toy is in the
container. This ability to distinguish self-knowledge from other-knowledge
allows the child to generate the prediction that the actor will search for her
toy in the original location. The ability to represent the fact that the actor
has a false belief about the location of the toy—a belief that differs from the
child’s own—is present by the second year of life, if not earlier (Baillargeon,
Scott, & He, 2010).
Perspective-taking is likewise a key contributor to adult interpersonal communication; effective speakers tailor how they speak to their audience, and
effective listeners adjust their understanding of what is said based on what
they know about their interlocutor (Clark, 1996). For example, when describing New York City landmarks, New Yorkers will use proper names (e.g.,
Rockefeller Center) when speaking to other New Yorkers and descriptive

Language, Perspective, and Memory

3

phrases (e.g., the building with the flags in front of it) when speaking to a
person who has never been to New York (Isaacs & Clark, 1987). Similarly, an
effective listener might interpret a word such as tweet with different meanings, depending on their beliefs about what the speaker knows. If a college
student were listening to a young person, tweet might bring to mind the social
media site, “Twitter”; if they were talking to an elderly person, the same word
might bring to mind a bird.
This ability to adjust our expectations about the actions of others and to tailor how we use language with other individuals critically depends on representations of the perspective of other individuals. In order to appreciate that
another person has a different perspective than one’s own, for example, individuals must store in memory a representation of their interlocutor’s knowledge state that is distinct from their own knowledge state. They must also
be able to access those memory representations quickly, when cued appropriately. The nature of these memory structures and the ways in which they
are accessed during language production and comprehension are areas of
active research. The products of this research have implications for a variety
of cognitive domains.
FOUNDATIONAL RESEARCH
Representations of the perspective of others play a central role in face-to-face
communicative settings. In conversation, the common ground between the
interlocutors is thought to be the fundamental backdrop against which communication takes place (Clark, 1996). As a result, many of the advances in
our understanding of how perspective is represented and used come from
psycholinguistic studies of how language is processed in face-to-face conversation. This research shows that both speakers and listeners alike tailor
language processes based on the knowledge that is jointly held with their
conversational partner.
COMMON GROUND IN LANGUAGE PRODUCTION
It is well known that when we refer to something in the world, we must
distinguish what we intend to describe from other potential referents (Olson,
1970; Osgood, 1971). When one orders a doughnut at a favorite doughnut
shop, it simply will not suffice to say “One doughnut, please.” Instead, the
speaker must specify her referential expression “one doughnut” to pick
out the one she wants, from many that she does not want, as in “One
glazed chocolate cake doughnut, please.” Thus, speakers design their referring
expressions with respect to the physical and cognitive environment they find
themselves in. Consistent with the logic that common ground is central to

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

language use (Clark, 1996), speakers also change their utterances depending
on their beliefs about what their conversational partner knows and does
not know. For example, imagine a situation in which a speaker will ask a
listener to hand her one of several items in a display, such as the smaller of
two triangles in Figure 2. In a situation where both the speaker and listener
see the two triangles, speakers typically specify the size of the triangle, as
in “Please hand me the small triangle”(Figure 2a). The adjective “small” plays
a critical role in this linguistic exchange, because it specifies which of the
two triangles the speaker is referring to. However, if the larger triangle
is not visible to their partner (Figure 2b), speakers often will refer to it
simply as “the triangle” because the contrasting adjective, “small,” becomes
superfluous (Nadig & Sedivy, 2002; Wardlow Lane, Groisman, & Ferreira,
2006) and potentially confusing. From the listener’s perspective, there is
only one triangle. Thus, speakers use their knowledge about the listener’s
perspective in the situation, in order to constrain what they say.
While the shared physical environment is one component of common
ground, another mechanism by which common ground is formed is through
communication itself. A central finding in studies of conversational language
is that, as we converse, the ways in which we refer to various topics of discussion become more definite and succinct over the course of a conversation
(Wilkes-Gibbs & Clark, 1992). For instance, when describing an ambiguous
black-and-white image such as the leftmost object in Figure 3, on a first
attempt, a speaker might say, “it resembles someone who looks like they’re trying
to climb stairs. There’s two feet, one is way above the other.” When redescribing
the same figure later in the conversation, the speaker might now refer to
it more concretely and efficiently, as in “the stair climber.” Memory for the
previous mention of that ambiguous figure, and the knowledge that this
memory is shared with one’s conversational partner, allows the speaker
to refine her phrasing. This phenomenon is observable even on a global
level. Consider that in the early days of the Internet, commercials enticed
potential customers to view their product on the “world wide web,” whereas
nowadays, we simply refer to the “web.” This shortening of the expression

Listener

(a)

Speaker

Listener

(b)

Speaker

Figure 2 (a) Both triangles are visible to the listener. (b) A barrier blocks the larger
triangle, making it hidden from the listener’s perspective.

Language, Perspective, and Memory

Figure 3

5

Example tangrams.

represents the growth in global common knowledge about the Internet, and,
as such, indicates that, as a global community, we have formed common
ground for this concept.
Key evidence in support of the argument that these phenomena represent
common ground between people (and not simply private knowledge, e.g.,
about what the web is) comes from findings that speakers no longer use these
shortened terms such as stair climber or web when speaking with a person who
is unknowledgeable about the term. In such circumstances, speakers revert
to longer, more descriptive phrases (Horton & Gerrig, 2002; Wilkes-Gibbs &
Clark, 1992). This ability to tailor referential expressions to the knowledge of
specific addressees is remarkably flexible and powerful. Speakers are able to
keep track of a large number of distinct and intermixed items, some of which
are shared with one partner, and some of which are shared with a different
partner (Horton & Gerrig, 2005). Speakers are also able to switch back and
forth between conversational partners with whom they share distinct knowledge, appropriately adjusting their expressions to be more descriptive when
the current addressee is “naïve” with respect to that given item, and using
established naming conventions when the current addressee is familiar with
the convention (Horton & Gerrig, 2002; Yoon & Brown-Schmidt, 2014).
COMMON GROUND IN LANGUAGE COMPREHENSION
In much the same way as speakers tailor how they speak depending on
the knowledge of their addressee, listeners, too, engage in complementary,
partner-specific adjustment during conversation. Once two people in a
conversation have converged on a particular expression (e.g., “the shiny
cylinder”) to refer to a given referent, listeners are surprised to hear the
same speaker switch to an equally plausible but novel expression (e.g., “the
silver pipe”), suggesting that the novel expression provides a less effective
cue to the representation of that referent stored in memory (Metzing &
Brennan, 2003). However, if a new person joins the conversation, these
novel expressions are no longer surprising, and are processed more quickly.

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

The fact that the penalty for novel expressions does not extend to speakers
who do not share common ground for the original term, shows that this
is not simply an egocentric process. Instead, listeners take into account
what knowledge their partner does and does not have when generating an
interpretation of what they say.
Listeners similarly use what they know about a speaker’s visual perspective
within the physical context when interpreting language (Hanna, Tanenhaus,
& Trueswell, 2003; Heller, Grodner, & Tanenhaus, 2008; Nadig & Sedivy,
2002). For instance, if a listener is looking at a display that includes two red
triangles, only one of which is visible to the speaker (Figure 4), when the
listener hears an instruction such as, “Click on the red triangle,” the listener
will look primarily at the red triangle in common ground. The listener is
able to rule out the triangle in privileged ground because she knows that,
from the speaker’s perspective, only one triangle is visible. Complementary
findings include evidence that if a speaker asks you to hand her something,
e.g., “Hand me the cake mix … ,” listeners interpret the speaker as requesting
the cake mix that is out of her own reach, and not a cake mix that she
could have reached from her position; after all, she would not need to ask
for it if it were within reach (Hanna & Tanenhaus, 2004). Further, when
interpreting informational questions, for example, “What’s above the … ,”
listeners readily interpret them as asking about something that the speaker
does not already know—otherwise, there is no need to ask a question

Figure 4 Listener’s view while listening to the instruction, “Click on the red
triangle.” The gray background indicates an object that is visible to the listener, but
not to the speaker.

Language, Perspective, and Memory

7

in the first place (Brown-Schmidt, 2012; Brown-Schmidt, Gunlogson, &
Tanenhaus, 2008). Taken together, this body of research shows that listeners
access representations of the speaker’s perspective, and incorporate this
information into the unfolding interpretation of the speaker’s utterances.
One way of conceptualizing this sensitivity to partner-specific common
ground is as a type of contextual integration, in which information, such as a
jointly established description of a referent (e.g., the shiny cylinder), is bound
to the person with whom the information was experienced. Framed in this
way, understanding how this binding of partner identity with information
from the conversation is encoded, maintained, and accessed in the service
of language processing becomes a central question. Research in the memory
tradition distinguishes concepts such as item memory, including memory
for a word on a list, from the source of that information, including the
room the words were studied in (see Johnson, Hashtroudi, & Lindsay,
1993, for a review). While source in the memory literature is conceptualized
broadly as the conditions in which information is encoded, in the domain of
conversational language, perhaps the most important source is the identity
and perspective of the person with whom you are conversing. Consistent
with this idea that interlocutors are particularly good contextual cues comes
from a meta-analysis of environmental context effects in memory research
(e.g., studies in which the locations of study and test are either the same
or different), which reported a large effect of whether the Experimenter
stayed the same from study to test (Smith & Vela, 2001). In what follows,
we discuss new research that furthers our understanding of the role of
context in a discourse and explores the neural underpinnings of discourse
representations. We then describe potential domains of inquiry that could
be pursued to fruitfully expand our understanding of the link between
memory and language as it pertains to perspective-taking.
CUTTING-EDGE RESEARCH
The literature on language processing in conversation has clearly established
that information in the local context guides language use, including visually
available information (Olson, 1970), as well as information about what
a partner does and does not see or know (Hanna et al., 2003; Nadig &
Sedivy, 2002). While it is clear that conversational partners encode and use
information about content that is discussed, such as the way particular
referents are described (Wilkes-Gibbs & Clark, 1992), less well established
is the extent to which contextual information, including information about
source, is encoded and maintained in conversation. Indeed, source memory
is often more fragile than item memory (e.g., Ferguson, Hashtroudi, &
Johnson, 1992).

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

An important question then is whether memory for source can be cued
effectively at later time points in a conversation, such that future referring is
altered as a result of previous contextual information, or whether the change
of temporal and cognitive context is enough to render the source information inaccessible. Moreover, what are the memory systems that support the
encoding and maintenance of item and source information to be accessed in
the discourse?
PAST CONTEXTS AND FUTURE LANGUAGE USE
According to the context-memory view of discourse (Yoon & Brown-Schmidt,
2013), speakers maintain memory for how they had described previous referents in the conversation. These memory representations guide not only how
speakers re-refer to these referents (Brennan & Clark, 1996; Wilkes-Gibbs &
Clark, 1992) but also how they describe other, previously unnamed referents
in the future (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013). Key evidence for persistent effects of past contexts on future referring comes from
experiments that examine how the way in which speakers refer to a given
referent in one context, shapes how the same item is referred to in a new context. For example, if a pair of conversational partners are presented with a
display that contains several different fish pictures (Figure 5a), and are asked
to refer to each picture, partners will establish brief but descriptive labels for
each fish (e.g., “the curved round fish”) that distinguish among the category
exemplars in the scene (i.e., the other two fish).
Evidence of persistent context effects comes from findings that even when
the context changes and now includes only a single category exemplar
(Figure 5b), speakers continue to use previously established expressions
(e.g., “the curved round fish”) even though they are overly specific given the
immediate context. In other words, while the abbreviated expression the fish

(a)

(b)

(c)

Figure 5 (a) Speakers say “the curved round fish” to disambiguate the target
(circled) fish from the other two fish. (b) Speakers say “the curved round fish” to
disambiguate the target fish from the contrasting fish seen previously (lexical
entrainment). (c) Speakers say “the glittery fish” to disambiguate the novel target
fish from previously discussed fish (lexical differentiation effect).

Language, Perspective, and Memory

9

would have uniquely identified the referent, speakers persist in using the
longer, more descriptive term that had been established previously (Brennan
& Clark, 1996). According to the context-memory view, the source of this
effect is that the speaker’s representation of the relevant context includes
both the single fish in the immediate visual context as well as multiple other
contrasting fish from past contexts.
Distinguishing this context-memory view from a simpler account on which
speakers had simply encoded the term the curved round fish as the name of
the referent (rather than encoding both the term and the items in the past
context), is supported by evidence that previous referents shape the descriptions of future referents. One demonstration of this is the phenomenon of
lexical differentiation (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), in
which speakers differentiate new discourse referents from similar referents
discussed in past contexts. For example, imagine a different situation in
which conversational partners first refer to the fish in Figure 5b as the fish. In
a subsequent context in which a different, novel fish is present (Figure 5c),
speakers are more likely to describe it with a modified noun phrase (e.g.,
glittery fish) compared to a situation where they had not previously described
other fish (Van der Wege, 2009). This lexical differentiation effect shows
that speakers consider how distinct referents from the discourse history
were previously described when designing new expressions. One intriguing
finding is that while Yoon and Brown-Schmidt (2013) observed lexical
differentiation in language production, there was no evidence that listeners
expected speakers to differentiate current from past referents. Instead,
listeners interpreted established terms such as fish quickly, regardless of
whether the term had previously been used to describe a different fish
or not. This suggests that the scope of the relevant discourse context for
speakers and listeners may differ, resulting in subtle asymmetries in the
relationship between reference and context.
MEMORY SYSTEMS IN LANGUAGE USE
The persistence of the historical discourse context in language use raises the
question of how this information is encoded in memory. At least some evidence suggests that multiple memory systems may be involved. In particular, neuropsychological work with individuals who have severe declarative
memory impairment (amnesia) due to bilateral hippocampal damage shows
patterns of deficits and sparing of function consistent with contributions of
multiple memory systems (Duff & Brown-Schmidt, 2012). The idea that the
hippocampus might be involved in processing partner-specific representations of common ground falls naturally from other findings pointing to hippocampal involvement in the processing of object–location relations, even

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

over short time-scales (Hannula, Tranel, & Cohen, 2006; Hannula & Ranganath, 2008). For example, individuals with amnesia are largely successful
at using information about what a conversational partner can and cannot see
in a visual display to constrain interpretation of referring expressions (Rubin,
Brown-Schmidt, Duff, Tranel, & Cohen, 2011). However, when the same individuals were presented with situations in which common ground was established linguistically (and no visual cues to what was shared and what was
not shared were present), they were no longer able to distinguish common
ground from privileged ground after a brief delay. Converging evidence from
studies of the ability to process short narratives reveals significant deficits in
the ability of individuals with amnesia to track discourse referents over the
course of a story, and use information about their relative salience to process pronouns that refer back to these characters (Kurczek, Brown-Schmidt, &
Duff, 2013). Yet, other evidence shows sparing of learning. Specifically, individuals with amnesia successfully develop talker-specific representations of
the accents of different talkers, and use these representations to guide speech
perception (Trude, Duff, & Brown-Schmidt, 2014). Together, these findings
emphasize the need to understand the memory systems that contribute to
and are necessary for the ability to tailor language use to one’s conversational
partner.
KEY ISSUES FOR FUTURE RESEARCH
The literature on language processing in conversation has clearly established
that better understanding the relationship between language and memory
processes is a key area in need of development, not only in terms of experimental findings but also in terms of theory building and the design of implemented models of the underlying processes. In what follows, we describe
several new and promising lines of inquiry that address these questions.
According to one recent proposal, the cognitive processes involved in
the design of a referring expression for another individual are analogous
to those in play when one is constructing a cue that will be used later to
access a specific event in memory (Tullis, 2013; Tullis & Benjamin, 2014).
In a cue-generation task, learners are given a large set of target words and
told to generate one or more cue words for each one. These cues are then
presented to the participants after a delay and they are told to retrieve the
target words they had previously read. Tullis found that learners generate
cues differently when they know that they will be used for future retrieval
compared to when they are simply asked to generate descriptions. They also
make their cues less distinctive and more strongly associated with the target
item when they are generating cues for another person compared to when
they are generating cues for themselves.

Language, Perspective, and Memory

11

The adjustments made during cue generation might reflect underlying processes similar to those occurring during perspective-taking in language use.
For instance, when taking notes in a class, the student’s goal is to jot down
short reminders that will, on later reading, trigger the recollection of the
larger, more complex piece of information that needs to be learned. It can
be challenging to create such reminders. Often they are not successful in eliciting the appropriate recollection (e.g., when the student rereads her notes
and cannot remember what they meant). In generating these cues, the student must essentially take the perspective of her future self and write notes
tailored to her future self’s perspective and knowledge state. Similarly, when
rereading notes, the student must use her knowledge of her past perspective
to constrain the possible interpretations of the cues.
An important question for future work is whether these two seemingly
analogous tasks—perspective-taking during language processing and during cue generation—do, in fact, reflect the same cognitive mechanisms, and,
if so, what the underlying domain-general cognitive functions might be. For
example, some work suggests that individuals who score higher on a test of
working memory are better at designing referring expressions based on the
perspective of their partner (Wardlow, 2013). If similar domain-general mechanisms such as working memory also confer specific benefits in effective cue
generation, this would make the prediction that individuals who are successful in one domain might also be successful in the other. This has important
implications for understanding the architecture of the cognitive mechanisms
underlying language processing and memory. Further, better understanding
how these fundamental skills are interconnected may lead to more strategic
interventions for those who may have deficits in either domain (e.g., older
adults).
Another important line of inquiry will be to identify how a participant’s
role in a conversation, for example, as speaker, listener, or overhearer,,
contributes to the encoding of conversational contextual cues. Recall that
Yoon and Brown-Schmidt (2013) found that while speakers showed sensitivity to the historical discourse context resulting in lexical differentiation
effects, listeners did not. This finding suggests that the representation of
the discourse record might differ for speakers and listeners. If so, critical
questions remain about how the way in which the historical discourse is
encoded in conversation—whether it is from a speaker’s perspective or a
listener’s—guides how this information is used. Conversational partners
are thought to encode contextually rich representations of joint experiences
(Clark & Marshall, 1978), possibly through automatic memory processes that
associate partners with experienced information (Horton & Gerrig, 2005),
and recent findings point to a tight degree of coupling in the representations
of speakers and listeners (Brown-Schmidt & Tanenhaus, 2008; Richardson,

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Dale, & Kirkham, 2007). Thus, understanding the extent to which coordinated representations are normative, and when representations are not
coordinated, would seem to be an important area of inquiry. Related to this
question is that of the breadth of the encoding of the discourse context. For
example, while it is clear that speakers maintain, and use, representations of
previous referents (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), an
open question is whether speakers consider other aspects of the previous
discourse context as well, such as unmentioned properties of previous
referents, or unmentioned objects in the discourse context. On the basis of
the findings in the memory tradition, the degree to which such contextual
information is encoded may depend on whether the encoding and test
conditions emphasize integration of the information with the context, or
instead whether associations among discussed items are more important
(see Eich, 1985; Smith & Vela, 2001).
As we move closer to the goal of developing theories of conversational
memory and language use, such efforts will undoubtedly benefit from
increased collaboration and interaction among researchers in both memoryand language-processing traditions. Research on conversation that combines
the visual world eye-tracking technique (Tanenhaus, Spivey-Knowlton,
Eberhard, & Sedivy, 1995), well established as a key method in language
processing research because it provides valuable information about the time
course of comprehension processes, with methods such as explicit memory
measures and signal detection analyses (Banks, 1970), will likely prove
invaluable in these efforts. In addition, the emerging trend of using Amazon’s Mechanical Turk to collect data about learning and language use over
time (e.g., Fine & Jaeger, 2013) will likely become an increasingly important
resource as it affords collecting large amounts of data from representative
populations. These new approaches, along with a more integrative view of
conversational language research, one that melds language and memory
perspectives (e.g., Stafford & Daly, 1984), will contribute to formulating a
more unified framework of these cognitive processes.

REFERENCES
Baillargeon, R., Scott, R. M., & He, Z. (2010). False-belief understanding in infants.
Trends in Cognitive Sciences, 14, 110–118.
Banks, W. P. (1970). Signal detection theory and human memory. Psychological Bulletin, 74, 81–99.
Brennan, S. E., & Clark, H. H. (1996). Conceptual pacts and lexical choice in conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22,
482–493.

Language, Perspective, and Memory

13

Brown-Schmidt, S. (2012). Beyond common and privileged: Gradient representations
of common ground in real-time language use. Language and Cognitive Processes, 27,
62–89.
Brown-Schmidt, S., Gunlogson, C., & Tanenhaus, M. K. (2008). Addressees distinguish shared from private information when interpreting questions during interactive conversation. Cognition, 107, 1122–1134.
Brown-Schmidt, S., & Tanenhaus, M. K. (2008). Real-time investigation of referential
domains in unscripted conversation: A targeted language game approach. Cognitive Science, 32, 643–684.
Clark, H. H. (1996). Using language. Cambridge, England: Cambridge University
Press.
Clark, H. H., & Marshall, C. R. (1978). Reference diaries. In D. L. Waltz (Ed.), Theoretical issues in natural language processing (Vol. 2, pp. 57–63). New York, NY: Association for Computing Machinery.
Duff, M. C., & Brown-Schmidt, S. (2012). The hippocampus and the flexible use and
processing of language. Frontiers in Cognitive Science, 6, 1–9.
Eich, E. (1985). Context, memory, and integrated item/context imagery. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 11, 764–770.
Ferguson, S. A., Hashtroudi, S., & Johnson, M. K. (1992). Age differences in using
source-relevant cues. Psychology and Aging, 7, 443.
Fine, A. B., & Jaeger, T. F. (2013). Syntactic priming in language comprehension
allows linguistic expectations to converge on the statistics of the input. In M.
Knauff, M. Pauen, N. Sebanz & I. Wachsmuth (Eds.), Proceedings of the 35th Annual
Meeting of the Cognitive Science Society (pp. 2279–2284). Austin, TX: Cognitive Science Society.
Hanna, J. E., & Tanenhaus, M. K. (2004). Pragmatic effects on reference resolution in
a collaborative task: Evidence from eye movements. Cognitive Science, 28, 105–115.
Hanna, J. E., Tanenhaus, M. K., & Trueswell, J. C. (2003). The effects of common
ground and perspective on domains of referential interpretation. Journal of Memory
and Language, 49, 43–61.
Hannula, D. E., & Ranganath, C. (2008). Medial temporal lobe activity predicts successful relational memory binding. The Journal of Neuroscience, 28, 116–124.
Hannula, D. E., Tranel, D., & Cohen, N. J. (2006). The long and the short of it: Relational memory impairments in amnesia, even at short lags. The Journal of Neuroscience, 26, 8352–8359.
Heller, D., Grodner, D., & Tanenhaus, M. K. (2008). The role of perspective in identifying domains of reference. Cognition, 108, 831–836.
Horton, W. S., & Gerrig, R. J. (2002). Speakers’ experiences and audience design:
Knowing when and knowing how to adjust utterances to addressees. Journal of
Memory and Language, 47, 589–606.
Horton, W. S., & Gerrig, R. J. (2005). The impact of memory demands on audience
design during language production. Cognition, 96, 127–142.
Isaacs, E. A., & Clark, H. H. (1987). References in conversation between experts and
novices. Journal of Experimental Psychology: General, 116, 26–37.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3.
Kurczek, J., Brown-Schmidt, S., & Duff, M. (2013). Hippocampal contributions to
language: Evidence of referential processing deficits in amnesia. Journal of Experimental Psychology: General, 142, 1346–1354.
Metzing, C., & Brennan, S. E. (2003). When conceptual pacts are broken: Partnerspecific effects on the comprehension of referring expressions. Journal of Memory
and Language, 49, 201–213.
Nadig, A. S., & Sedivy, J. C. (2002). Evidence of perspective-taking constraints in
children’s on-line reference resolution. Psychological Science, 13, 329–336.
Olson, D. R. (1970). Language and thought: Aspects of a cognitive theory of semantics. Psychological Review, 77, 257–273.
Osgood, C. E. (1971). Exploration in semantic space: A personal diary. Journal of Social
Issues, 27, 5–64.
Richardson, D. C., Dale, R., & Kirkham, N. Z. (2007). The art of conversation is coordination common ground and the coupling of eye movements during dialogue.
Psychological Science, 18, 407–413.
Rubin, R. D., Brown-Schmidt, S., Duff, M. C., Tranel, D., & Cohen, N. J. (2011).
How do I remember that I know you know that I know? Psychological Science, 22,
1574–1582.
Smith, S. M., & Vela, E. (2001). Environmental context-dependent memory: A review
and meta-analysis. Psychonomic Bulletin & Review, 8, 203–220.
Stafford, L., & Daly, J. A. (1984). Conversational memory. Human Communication
Research, 10, 379–402.
Tanenhaus, M. K., Spivey-Knowlton, M. J., Eberhard, K. M., & Sedivy, J. E. (1995).
Integration of visual and linguistic information in spoken language comprehension. Science, 268, 1632–1634.
Trude, A. M., Duff, M., & Brown-Schmidt, S. (2014). Talker-specific learning in amnesia: Insight into mechanisms of adaptive speech perception. Cortex, 54, 117–123.
Tullis, J. G. (2013). Cue generation: How learners flexibly support future retrieval (Unpublished doctoral dissertation). University of Illinois at Urbana-Champaign, Illinois.
Tullis, J. G., & Benjamin, A. S. (in press). Cueing others’ memories. Memory & Cognition.
Van Der Wege, M. M. (2009). Lexical entrainment and lexical differentiation in reference phrase choice. Journal of Memory and Language, 60, 448–463.
Wardlow, L. (2013). Individual differences in speakers’ perspective taking: The roles
of executive control and working memory. Psychonomic Bulletin & Review, 20,
766–772.
Wardlow Lane, L., Groisman, M., & Ferreira, V. S. (2006). Don’t talk about pink
elephants! Speaker’s control over leaking private information during language
production. Psychological Science, 17, 273–277.
Wilkes-Gibbs, D., & Clark, H. H. (1992). Coordinating beliefs in conversation. Journal
of Memory and Language, 31, 183–194.
Yoon, S. O., & Brown-Schmidt, S. (2013). Lexical differentiation in language production and comprehension. Journal of Memory and Language, 69, 397–416.

Language, Perspective, and Memory

15

Yoon, S. O., & Brown-Schmidt, S. (2014). Adjusting conceptual pacts in three-party
conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition,
40, 919–937.

RACHEL A. RYSKIN SHORT BIOGRAPHY
Rachel A. Ryskin is a graduate student in the Department of Psychology at
the University of Illinois at Urbana Champaign, rryskin@gmail.com.
SI ON YOON SHORT BIOGRAPHY
Si On Yoon is a graduate student in the Department of Psychology at the
University of Illinois at Urbana Champaign, sion0912@gmail.com.
SARAH BROWN-SCHMIDT SHORT BIOGRAPHY
Sarah Brown-Schmidt is an Assistant Professor in the Department of
Psychology at the University of Illinois at Urbana Champaign, sarahbrownschmidt@gmail.com. Webpage: sarahbrownschmidt.com/professional.
RELATED ESSAYS
Theory of Mind and Behavior (Psychology), Amanda C. Brandone
Delusions (Psychology), Max Coltheart
Misinformation and How to Correct It (Psychology), John Cook et al.
Insight (Psychology), Brian Erickson and John Kounios
Cognitive Processes Involved in Stereotyping (Psychology), Susan T. Fiske and
Cydney H. Dupree
Language and Thought (Psychology), Susan Goldin-Meadow
Concepts and Semantic Memory (Psychology), Barbara C. Malt
Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Attention and Perception (Psychology), Ronald A. Rensink
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
Speech Perception (Psychology), Athena Vouloumanos
Theory of Mind (Psychology), Henry Wellman


Language, Perspective, and Memory
RACHEL A. RYSKIN, SI ON YOON, and SARAH BROWN-SCHMIDT

Abstract
The ability to take the perspective of another person is ubiquitous in many everyday
cognitive activities. In particular, it allows people to communicate efficiently with
conversational partners. Speakers tailor what they say based on the listener’s knowledge and, likewise, listeners use what they know about the speaker to better understand what the speaker means. In this essay, we review foundational research on the
role of perspective-taking in the domain of language processing and describe new
lines of work that are beginning to explore the memory processes that support the
efficient use of perspectives in conversation. We then discuss key avenues for future
research, such as investigating whether the type of perspective-taking involved in
creating memory reminders draws on the same underlying cognitive processes as
in the domain of language processing. Exploring this interface between language,
perspective-taking, and memory will require interdisciplinary crosstalk and integration of methodologies across the domains of memory and language research.

INTRODUCTION
Perspective-taking is the ability to mentally represent the beliefs and knowledge of another person, which may or may not differ from one’s own. More
generally, the capacity to appreciate alternative perspectives is a key contributor to a variety of human cognitive activities—from being able to read a
map (Figure 1) to interacting with another person. Consider that when any
two individuals come together, they have a certain amount of knowledge
about the world that is shared between them. For two Americans sitting in
a ball-park watching a game, this joint knowledge or common ground would
include popular cultural references such as the name of the President of the
United States, as well as immediately available information in the context
such as information about the last play, or the score in the game. These individuals would also have a certain amount of information that is private to
each individual, and not shared with the other person, for example, what
each one of them ate for breakfast. A central goal of many interactions is
sharing some of this private knowledge, sometimes called privileged ground, in
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.

1

2

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

School

Home

A

B
Toy store

Figure 1 A and B are both reading the same map from opposite perspectives.
For A, the school is on the left. For B, the school is on the right. If B is directing A
about how to walk from home to the school, B will have to take the opposite
perspective (or vice versa); otherwise, A might end up at the toy store.

order to introduce new information and keep the conversation interesting
(after all, a conversation is bound to be painfully boring if it only involved
repeating to one another what is already jointly known).
Keeping track of what others do and do not know is an ability that is
observed even in very young children. For example, when young children
see an actor place a toy inside a container and leave the room, they expect
the actor, on her return, to search for the toy in that same container even
if the child knows that the toy was removed from this container. Thus, the
child is able to distinguish their privileged knowledge that the toy is no
longer in the container, from the actor’s false belief that the toy is in the
container. This ability to distinguish self-knowledge from other-knowledge
allows the child to generate the prediction that the actor will search for her
toy in the original location. The ability to represent the fact that the actor
has a false belief about the location of the toy—a belief that differs from the
child’s own—is present by the second year of life, if not earlier (Baillargeon,
Scott, & He, 2010).
Perspective-taking is likewise a key contributor to adult interpersonal communication; effective speakers tailor how they speak to their audience, and
effective listeners adjust their understanding of what is said based on what
they know about their interlocutor (Clark, 1996). For example, when describing New York City landmarks, New Yorkers will use proper names (e.g.,
Rockefeller Center) when speaking to other New Yorkers and descriptive

Language, Perspective, and Memory

3

phrases (e.g., the building with the flags in front of it) when speaking to a
person who has never been to New York (Isaacs & Clark, 1987). Similarly, an
effective listener might interpret a word such as tweet with different meanings, depending on their beliefs about what the speaker knows. If a college
student were listening to a young person, tweet might bring to mind the social
media site, “Twitter”; if they were talking to an elderly person, the same word
might bring to mind a bird.
This ability to adjust our expectations about the actions of others and to tailor how we use language with other individuals critically depends on representations of the perspective of other individuals. In order to appreciate that
another person has a different perspective than one’s own, for example, individuals must store in memory a representation of their interlocutor’s knowledge state that is distinct from their own knowledge state. They must also
be able to access those memory representations quickly, when cued appropriately. The nature of these memory structures and the ways in which they
are accessed during language production and comprehension are areas of
active research. The products of this research have implications for a variety
of cognitive domains.
FOUNDATIONAL RESEARCH
Representations of the perspective of others play a central role in face-to-face
communicative settings. In conversation, the common ground between the
interlocutors is thought to be the fundamental backdrop against which communication takes place (Clark, 1996). As a result, many of the advances in
our understanding of how perspective is represented and used come from
psycholinguistic studies of how language is processed in face-to-face conversation. This research shows that both speakers and listeners alike tailor
language processes based on the knowledge that is jointly held with their
conversational partner.
COMMON GROUND IN LANGUAGE PRODUCTION
It is well known that when we refer to something in the world, we must
distinguish what we intend to describe from other potential referents (Olson,
1970; Osgood, 1971). When one orders a doughnut at a favorite doughnut
shop, it simply will not suffice to say “One doughnut, please.” Instead, the
speaker must specify her referential expression “one doughnut” to pick
out the one she wants, from many that she does not want, as in “One
glazed chocolate cake doughnut, please.” Thus, speakers design their referring
expressions with respect to the physical and cognitive environment they find
themselves in. Consistent with the logic that common ground is central to

4

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

language use (Clark, 1996), speakers also change their utterances depending
on their beliefs about what their conversational partner knows and does
not know. For example, imagine a situation in which a speaker will ask a
listener to hand her one of several items in a display, such as the smaller of
two triangles in Figure 2. In a situation where both the speaker and listener
see the two triangles, speakers typically specify the size of the triangle, as
in “Please hand me the small triangle”(Figure 2a). The adjective “small” plays
a critical role in this linguistic exchange, because it specifies which of the
two triangles the speaker is referring to. However, if the larger triangle
is not visible to their partner (Figure 2b), speakers often will refer to it
simply as “the triangle” because the contrasting adjective, “small,” becomes
superfluous (Nadig & Sedivy, 2002; Wardlow Lane, Groisman, & Ferreira,
2006) and potentially confusing. From the listener’s perspective, there is
only one triangle. Thus, speakers use their knowledge about the listener’s
perspective in the situation, in order to constrain what they say.
While the shared physical environment is one component of common
ground, another mechanism by which common ground is formed is through
communication itself. A central finding in studies of conversational language
is that, as we converse, the ways in which we refer to various topics of discussion become more definite and succinct over the course of a conversation
(Wilkes-Gibbs & Clark, 1992). For instance, when describing an ambiguous
black-and-white image such as the leftmost object in Figure 3, on a first
attempt, a speaker might say, “it resembles someone who looks like they’re trying
to climb stairs. There’s two feet, one is way above the other.” When redescribing
the same figure later in the conversation, the speaker might now refer to
it more concretely and efficiently, as in “the stair climber.” Memory for the
previous mention of that ambiguous figure, and the knowledge that this
memory is shared with one’s conversational partner, allows the speaker
to refine her phrasing. This phenomenon is observable even on a global
level. Consider that in the early days of the Internet, commercials enticed
potential customers to view their product on the “world wide web,” whereas
nowadays, we simply refer to the “web.” This shortening of the expression

Listener

(a)

Speaker

Listener

(b)

Speaker

Figure 2 (a) Both triangles are visible to the listener. (b) A barrier blocks the larger
triangle, making it hidden from the listener’s perspective.

Language, Perspective, and Memory

Figure 3

5

Example tangrams.

represents the growth in global common knowledge about the Internet, and,
as such, indicates that, as a global community, we have formed common
ground for this concept.
Key evidence in support of the argument that these phenomena represent
common ground between people (and not simply private knowledge, e.g.,
about what the web is) comes from findings that speakers no longer use these
shortened terms such as stair climber or web when speaking with a person who
is unknowledgeable about the term. In such circumstances, speakers revert
to longer, more descriptive phrases (Horton & Gerrig, 2002; Wilkes-Gibbs &
Clark, 1992). This ability to tailor referential expressions to the knowledge of
specific addressees is remarkably flexible and powerful. Speakers are able to
keep track of a large number of distinct and intermixed items, some of which
are shared with one partner, and some of which are shared with a different
partner (Horton & Gerrig, 2005). Speakers are also able to switch back and
forth between conversational partners with whom they share distinct knowledge, appropriately adjusting their expressions to be more descriptive when
the current addressee is “naïve” with respect to that given item, and using
established naming conventions when the current addressee is familiar with
the convention (Horton & Gerrig, 2002; Yoon & Brown-Schmidt, 2014).
COMMON GROUND IN LANGUAGE COMPREHENSION
In much the same way as speakers tailor how they speak depending on
the knowledge of their addressee, listeners, too, engage in complementary,
partner-specific adjustment during conversation. Once two people in a
conversation have converged on a particular expression (e.g., “the shiny
cylinder”) to refer to a given referent, listeners are surprised to hear the
same speaker switch to an equally plausible but novel expression (e.g., “the
silver pipe”), suggesting that the novel expression provides a less effective
cue to the representation of that referent stored in memory (Metzing &
Brennan, 2003). However, if a new person joins the conversation, these
novel expressions are no longer surprising, and are processed more quickly.

6

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

The fact that the penalty for novel expressions does not extend to speakers
who do not share common ground for the original term, shows that this
is not simply an egocentric process. Instead, listeners take into account
what knowledge their partner does and does not have when generating an
interpretation of what they say.
Listeners similarly use what they know about a speaker’s visual perspective
within the physical context when interpreting language (Hanna, Tanenhaus,
& Trueswell, 2003; Heller, Grodner, & Tanenhaus, 2008; Nadig & Sedivy,
2002). For instance, if a listener is looking at a display that includes two red
triangles, only one of which is visible to the speaker (Figure 4), when the
listener hears an instruction such as, “Click on the red triangle,” the listener
will look primarily at the red triangle in common ground. The listener is
able to rule out the triangle in privileged ground because she knows that,
from the speaker’s perspective, only one triangle is visible. Complementary
findings include evidence that if a speaker asks you to hand her something,
e.g., “Hand me the cake mix … ,” listeners interpret the speaker as requesting
the cake mix that is out of her own reach, and not a cake mix that she
could have reached from her position; after all, she would not need to ask
for it if it were within reach (Hanna & Tanenhaus, 2004). Further, when
interpreting informational questions, for example, “What’s above the … ,”
listeners readily interpret them as asking about something that the speaker
does not already know—otherwise, there is no need to ask a question

Figure 4 Listener’s view while listening to the instruction, “Click on the red
triangle.” The gray background indicates an object that is visible to the listener, but
not to the speaker.

Language, Perspective, and Memory

7

in the first place (Brown-Schmidt, 2012; Brown-Schmidt, Gunlogson, &
Tanenhaus, 2008). Taken together, this body of research shows that listeners
access representations of the speaker’s perspective, and incorporate this
information into the unfolding interpretation of the speaker’s utterances.
One way of conceptualizing this sensitivity to partner-specific common
ground is as a type of contextual integration, in which information, such as a
jointly established description of a referent (e.g., the shiny cylinder), is bound
to the person with whom the information was experienced. Framed in this
way, understanding how this binding of partner identity with information
from the conversation is encoded, maintained, and accessed in the service
of language processing becomes a central question. Research in the memory
tradition distinguishes concepts such as item memory, including memory
for a word on a list, from the source of that information, including the
room the words were studied in (see Johnson, Hashtroudi, & Lindsay,
1993, for a review). While source in the memory literature is conceptualized
broadly as the conditions in which information is encoded, in the domain of
conversational language, perhaps the most important source is the identity
and perspective of the person with whom you are conversing. Consistent
with this idea that interlocutors are particularly good contextual cues comes
from a meta-analysis of environmental context effects in memory research
(e.g., studies in which the locations of study and test are either the same
or different), which reported a large effect of whether the Experimenter
stayed the same from study to test (Smith & Vela, 2001). In what follows,
we discuss new research that furthers our understanding of the role of
context in a discourse and explores the neural underpinnings of discourse
representations. We then describe potential domains of inquiry that could
be pursued to fruitfully expand our understanding of the link between
memory and language as it pertains to perspective-taking.
CUTTING-EDGE RESEARCH
The literature on language processing in conversation has clearly established
that information in the local context guides language use, including visually
available information (Olson, 1970), as well as information about what
a partner does and does not see or know (Hanna et al., 2003; Nadig &
Sedivy, 2002). While it is clear that conversational partners encode and use
information about content that is discussed, such as the way particular
referents are described (Wilkes-Gibbs & Clark, 1992), less well established
is the extent to which contextual information, including information about
source, is encoded and maintained in conversation. Indeed, source memory
is often more fragile than item memory (e.g., Ferguson, Hashtroudi, &
Johnson, 1992).

8

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

An important question then is whether memory for source can be cued
effectively at later time points in a conversation, such that future referring is
altered as a result of previous contextual information, or whether the change
of temporal and cognitive context is enough to render the source information inaccessible. Moreover, what are the memory systems that support the
encoding and maintenance of item and source information to be accessed in
the discourse?
PAST CONTEXTS AND FUTURE LANGUAGE USE
According to the context-memory view of discourse (Yoon & Brown-Schmidt,
2013), speakers maintain memory for how they had described previous referents in the conversation. These memory representations guide not only how
speakers re-refer to these referents (Brennan & Clark, 1996; Wilkes-Gibbs &
Clark, 1992) but also how they describe other, previously unnamed referents
in the future (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013). Key evidence for persistent effects of past contexts on future referring comes from
experiments that examine how the way in which speakers refer to a given
referent in one context, shapes how the same item is referred to in a new context. For example, if a pair of conversational partners are presented with a
display that contains several different fish pictures (Figure 5a), and are asked
to refer to each picture, partners will establish brief but descriptive labels for
each fish (e.g., “the curved round fish”) that distinguish among the category
exemplars in the scene (i.e., the other two fish).
Evidence of persistent context effects comes from findings that even when
the context changes and now includes only a single category exemplar
(Figure 5b), speakers continue to use previously established expressions
(e.g., “the curved round fish”) even though they are overly specific given the
immediate context. In other words, while the abbreviated expression the fish

(a)

(b)

(c)

Figure 5 (a) Speakers say “the curved round fish” to disambiguate the target
(circled) fish from the other two fish. (b) Speakers say “the curved round fish” to
disambiguate the target fish from the contrasting fish seen previously (lexical
entrainment). (c) Speakers say “the glittery fish” to disambiguate the novel target
fish from previously discussed fish (lexical differentiation effect).

Language, Perspective, and Memory

9

would have uniquely identified the referent, speakers persist in using the
longer, more descriptive term that had been established previously (Brennan
& Clark, 1996). According to the context-memory view, the source of this
effect is that the speaker’s representation of the relevant context includes
both the single fish in the immediate visual context as well as multiple other
contrasting fish from past contexts.
Distinguishing this context-memory view from a simpler account on which
speakers had simply encoded the term the curved round fish as the name of
the referent (rather than encoding both the term and the items in the past
context), is supported by evidence that previous referents shape the descriptions of future referents. One demonstration of this is the phenomenon of
lexical differentiation (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), in
which speakers differentiate new discourse referents from similar referents
discussed in past contexts. For example, imagine a different situation in
which conversational partners first refer to the fish in Figure 5b as the fish. In
a subsequent context in which a different, novel fish is present (Figure 5c),
speakers are more likely to describe it with a modified noun phrase (e.g.,
glittery fish) compared to a situation where they had not previously described
other fish (Van der Wege, 2009). This lexical differentiation effect shows
that speakers consider how distinct referents from the discourse history
were previously described when designing new expressions. One intriguing
finding is that while Yoon and Brown-Schmidt (2013) observed lexical
differentiation in language production, there was no evidence that listeners
expected speakers to differentiate current from past referents. Instead,
listeners interpreted established terms such as fish quickly, regardless of
whether the term had previously been used to describe a different fish
or not. This suggests that the scope of the relevant discourse context for
speakers and listeners may differ, resulting in subtle asymmetries in the
relationship between reference and context.
MEMORY SYSTEMS IN LANGUAGE USE
The persistence of the historical discourse context in language use raises the
question of how this information is encoded in memory. At least some evidence suggests that multiple memory systems may be involved. In particular, neuropsychological work with individuals who have severe declarative
memory impairment (amnesia) due to bilateral hippocampal damage shows
patterns of deficits and sparing of function consistent with contributions of
multiple memory systems (Duff & Brown-Schmidt, 2012). The idea that the
hippocampus might be involved in processing partner-specific representations of common ground falls naturally from other findings pointing to hippocampal involvement in the processing of object–location relations, even

10

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

over short time-scales (Hannula, Tranel, & Cohen, 2006; Hannula & Ranganath, 2008). For example, individuals with amnesia are largely successful
at using information about what a conversational partner can and cannot see
in a visual display to constrain interpretation of referring expressions (Rubin,
Brown-Schmidt, Duff, Tranel, & Cohen, 2011). However, when the same individuals were presented with situations in which common ground was established linguistically (and no visual cues to what was shared and what was
not shared were present), they were no longer able to distinguish common
ground from privileged ground after a brief delay. Converging evidence from
studies of the ability to process short narratives reveals significant deficits in
the ability of individuals with amnesia to track discourse referents over the
course of a story, and use information about their relative salience to process pronouns that refer back to these characters (Kurczek, Brown-Schmidt, &
Duff, 2013). Yet, other evidence shows sparing of learning. Specifically, individuals with amnesia successfully develop talker-specific representations of
the accents of different talkers, and use these representations to guide speech
perception (Trude, Duff, & Brown-Schmidt, 2014). Together, these findings
emphasize the need to understand the memory systems that contribute to
and are necessary for the ability to tailor language use to one’s conversational
partner.
KEY ISSUES FOR FUTURE RESEARCH
The literature on language processing in conversation has clearly established
that better understanding the relationship between language and memory
processes is a key area in need of development, not only in terms of experimental findings but also in terms of theory building and the design of implemented models of the underlying processes. In what follows, we describe
several new and promising lines of inquiry that address these questions.
According to one recent proposal, the cognitive processes involved in
the design of a referring expression for another individual are analogous
to those in play when one is constructing a cue that will be used later to
access a specific event in memory (Tullis, 2013; Tullis & Benjamin, 2014).
In a cue-generation task, learners are given a large set of target words and
told to generate one or more cue words for each one. These cues are then
presented to the participants after a delay and they are told to retrieve the
target words they had previously read. Tullis found that learners generate
cues differently when they know that they will be used for future retrieval
compared to when they are simply asked to generate descriptions. They also
make their cues less distinctive and more strongly associated with the target
item when they are generating cues for another person compared to when
they are generating cues for themselves.

Language, Perspective, and Memory

11

The adjustments made during cue generation might reflect underlying processes similar to those occurring during perspective-taking in language use.
For instance, when taking notes in a class, the student’s goal is to jot down
short reminders that will, on later reading, trigger the recollection of the
larger, more complex piece of information that needs to be learned. It can
be challenging to create such reminders. Often they are not successful in eliciting the appropriate recollection (e.g., when the student rereads her notes
and cannot remember what they meant). In generating these cues, the student must essentially take the perspective of her future self and write notes
tailored to her future self’s perspective and knowledge state. Similarly, when
rereading notes, the student must use her knowledge of her past perspective
to constrain the possible interpretations of the cues.
An important question for future work is whether these two seemingly
analogous tasks—perspective-taking during language processing and during cue generation—do, in fact, reflect the same cognitive mechanisms, and,
if so, what the underlying domain-general cognitive functions might be. For
example, some work suggests that individuals who score higher on a test of
working memory are better at designing referring expressions based on the
perspective of their partner (Wardlow, 2013). If similar domain-general mechanisms such as working memory also confer specific benefits in effective cue
generation, this would make the prediction that individuals who are successful in one domain might also be successful in the other. This has important
implications for understanding the architecture of the cognitive mechanisms
underlying language processing and memory. Further, better understanding
how these fundamental skills are interconnected may lead to more strategic
interventions for those who may have deficits in either domain (e.g., older
adults).
Another important line of inquiry will be to identify how a participant’s
role in a conversation, for example, as speaker, listener, or overhearer,,
contributes to the encoding of conversational contextual cues. Recall that
Yoon and Brown-Schmidt (2013) found that while speakers showed sensitivity to the historical discourse context resulting in lexical differentiation
effects, listeners did not. This finding suggests that the representation of
the discourse record might differ for speakers and listeners. If so, critical
questions remain about how the way in which the historical discourse is
encoded in conversation—whether it is from a speaker’s perspective or a
listener’s—guides how this information is used. Conversational partners
are thought to encode contextually rich representations of joint experiences
(Clark & Marshall, 1978), possibly through automatic memory processes that
associate partners with experienced information (Horton & Gerrig, 2005),
and recent findings point to a tight degree of coupling in the representations
of speakers and listeners (Brown-Schmidt & Tanenhaus, 2008; Richardson,

12

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Dale, & Kirkham, 2007). Thus, understanding the extent to which coordinated representations are normative, and when representations are not
coordinated, would seem to be an important area of inquiry. Related to this
question is that of the breadth of the encoding of the discourse context. For
example, while it is clear that speakers maintain, and use, representations of
previous referents (Van der Wege, 2009; Yoon & Brown-Schmidt, 2013), an
open question is whether speakers consider other aspects of the previous
discourse context as well, such as unmentioned properties of previous
referents, or unmentioned objects in the discourse context. On the basis of
the findings in the memory tradition, the degree to which such contextual
information is encoded may depend on whether the encoding and test
conditions emphasize integration of the information with the context, or
instead whether associations among discussed items are more important
(see Eich, 1985; Smith & Vela, 2001).
As we move closer to the goal of developing theories of conversational
memory and language use, such efforts will undoubtedly benefit from
increased collaboration and interaction among researchers in both memoryand language-processing traditions. Research on conversation that combines
the visual world eye-tracking technique (Tanenhaus, Spivey-Knowlton,
Eberhard, & Sedivy, 1995), well established as a key method in language
processing research because it provides valuable information about the time
course of comprehension processes, with methods such as explicit memory
measures and signal detection analyses (Banks, 1970), will likely prove
invaluable in these efforts. In addition, the emerging trend of using Amazon’s Mechanical Turk to collect data about learning and language use over
time (e.g., Fine & Jaeger, 2013) will likely become an increasingly important
resource as it affords collecting large amounts of data from representative
populations. These new approaches, along with a more integrative view of
conversational language research, one that melds language and memory
perspectives (e.g., Stafford & Daly, 1984), will contribute to formulating a
more unified framework of these cognitive processes.

REFERENCES
Baillargeon, R., Scott, R. M., & He, Z. (2010). False-belief understanding in infants.
Trends in Cognitive Sciences, 14, 110–118.
Banks, W. P. (1970). Signal detection theory and human memory. Psychological Bulletin, 74, 81–99.
Brennan, S. E., & Clark, H. H. (1996). Conceptual pacts and lexical choice in conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22,
482–493.

Language, Perspective, and Memory

13

Brown-Schmidt, S. (2012). Beyond common and privileged: Gradient representations
of common ground in real-time language use. Language and Cognitive Processes, 27,
62–89.
Brown-Schmidt, S., Gunlogson, C., & Tanenhaus, M. K. (2008). Addressees distinguish shared from private information when interpreting questions during interactive conversation. Cognition, 107, 1122–1134.
Brown-Schmidt, S., & Tanenhaus, M. K. (2008). Real-time investigation of referential
domains in unscripted conversation: A targeted language game approach. Cognitive Science, 32, 643–684.
Clark, H. H. (1996). Using language. Cambridge, England: Cambridge University
Press.
Clark, H. H., & Marshall, C. R. (1978). Reference diaries. In D. L. Waltz (Ed.), Theoretical issues in natural language processing (Vol. 2, pp. 57–63). New York, NY: Association for Computing Machinery.
Duff, M. C., & Brown-Schmidt, S. (2012). The hippocampus and the flexible use and
processing of language. Frontiers in Cognitive Science, 6, 1–9.
Eich, E. (1985). Context, memory, and integrated item/context imagery. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 11, 764–770.
Ferguson, S. A., Hashtroudi, S., & Johnson, M. K. (1992). Age differences in using
source-relevant cues. Psychology and Aging, 7, 443.
Fine, A. B., & Jaeger, T. F. (2013). Syntactic priming in language comprehension
allows linguistic expectations to converge on the statistics of the input. In M.
Knauff, M. Pauen, N. Sebanz & I. Wachsmuth (Eds.), Proceedings of the 35th Annual
Meeting of the Cognitive Science Society (pp. 2279–2284). Austin, TX: Cognitive Science Society.
Hanna, J. E., & Tanenhaus, M. K. (2004). Pragmatic effects on reference resolution in
a collaborative task: Evidence from eye movements. Cognitive Science, 28, 105–115.
Hanna, J. E., Tanenhaus, M. K., & Trueswell, J. C. (2003). The effects of common
ground and perspective on domains of referential interpretation. Journal of Memory
and Language, 49, 43–61.
Hannula, D. E., & Ranganath, C. (2008). Medial temporal lobe activity predicts successful relational memory binding. The Journal of Neuroscience, 28, 116–124.
Hannula, D. E., Tranel, D., & Cohen, N. J. (2006). The long and the short of it: Relational memory impairments in amnesia, even at short lags. The Journal of Neuroscience, 26, 8352–8359.
Heller, D., Grodner, D., & Tanenhaus, M. K. (2008). The role of perspective in identifying domains of reference. Cognition, 108, 831–836.
Horton, W. S., & Gerrig, R. J. (2002). Speakers’ experiences and audience design:
Knowing when and knowing how to adjust utterances to addressees. Journal of
Memory and Language, 47, 589–606.
Horton, W. S., & Gerrig, R. J. (2005). The impact of memory demands on audience
design during language production. Cognition, 96, 127–142.
Isaacs, E. A., & Clark, H. H. (1987). References in conversation between experts and
novices. Journal of Experimental Psychology: General, 116, 26–37.

14

EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES

Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 3.
Kurczek, J., Brown-Schmidt, S., & Duff, M. (2013). Hippocampal contributions to
language: Evidence of referential processing deficits in amnesia. Journal of Experimental Psychology: General, 142, 1346–1354.
Metzing, C., & Brennan, S. E. (2003). When conceptual pacts are broken: Partnerspecific effects on the comprehension of referring expressions. Journal of Memory
and Language, 49, 201–213.
Nadig, A. S., & Sedivy, J. C. (2002). Evidence of perspective-taking constraints in
children’s on-line reference resolution. Psychological Science, 13, 329–336.
Olson, D. R. (1970). Language and thought: Aspects of a cognitive theory of semantics. Psychological Review, 77, 257–273.
Osgood, C. E. (1971). Exploration in semantic space: A personal diary. Journal of Social
Issues, 27, 5–64.
Richardson, D. C., Dale, R., & Kirkham, N. Z. (2007). The art of conversation is coordination common ground and the coupling of eye movements during dialogue.
Psychological Science, 18, 407–413.
Rubin, R. D., Brown-Schmidt, S., Duff, M. C., Tranel, D., & Cohen, N. J. (2011).
How do I remember that I know you know that I know? Psychological Science, 22,
1574–1582.
Smith, S. M., & Vela, E. (2001). Environmental context-dependent memory: A review
and meta-analysis. Psychonomic Bulletin & Review, 8, 203–220.
Stafford, L., & Daly, J. A. (1984). Conversational memory. Human Communication
Research, 10, 379–402.
Tanenhaus, M. K., Spivey-Knowlton, M. J., Eberhard, K. M., & Sedivy, J. E. (1995).
Integration of visual and linguistic information in spoken language comprehension. Science, 268, 1632–1634.
Trude, A. M., Duff, M., & Brown-Schmidt, S. (2014). Talker-specific learning in amnesia: Insight into mechanisms of adaptive speech perception. Cortex, 54, 117–123.
Tullis, J. G. (2013). Cue generation: How learners flexibly support future retrieval (Unpublished doctoral dissertation). University of Illinois at Urbana-Champaign, Illinois.
Tullis, J. G., & Benjamin, A. S. (in press). Cueing others’ memories. Memory & Cognition.
Van Der Wege, M. M. (2009). Lexical entrainment and lexical differentiation in reference phrase choice. Journal of Memory and Language, 60, 448–463.
Wardlow, L. (2013). Individual differences in speakers’ perspective taking: The roles
of executive control and working memory. Psychonomic Bulletin & Review, 20,
766–772.
Wardlow Lane, L., Groisman, M., & Ferreira, V. S. (2006). Don’t talk about pink
elephants! Speaker’s control over leaking private information during language
production. Psychological Science, 17, 273–277.
Wilkes-Gibbs, D., & Clark, H. H. (1992). Coordinating beliefs in conversation. Journal
of Memory and Language, 31, 183–194.
Yoon, S. O., & Brown-Schmidt, S. (2013). Lexical differentiation in language production and comprehension. Journal of Memory and Language, 69, 397–416.

Language, Perspective, and Memory

15

Yoon, S. O., & Brown-Schmidt, S. (2014). Adjusting conceptual pacts in three-party
conversation. Journal of Experimental Psychology: Learning, Memory, and Cognition,
40, 919–937.

RACHEL A. RYSKIN SHORT BIOGRAPHY
Rachel A. Ryskin is a graduate student in the Department of Psychology at
the University of Illinois at Urbana Champaign, rryskin@gmail.com.
SI ON YOON SHORT BIOGRAPHY
Si On Yoon is a graduate student in the Department of Psychology at the
University of Illinois at Urbana Champaign, sion0912@gmail.com.
SARAH BROWN-SCHMIDT SHORT BIOGRAPHY
Sarah Brown-Schmidt is an Assistant Professor in the Department of
Psychology at the University of Illinois at Urbana Champaign, sarahbrownschmidt@gmail.com. Webpage: sarahbrownschmidt.com/professional.
RELATED ESSAYS
Theory of Mind and Behavior (Psychology), Amanda C. Brandone
Delusions (Psychology), Max Coltheart
Misinformation and How to Correct It (Psychology), John Cook et al.
Insight (Psychology), Brian Erickson and John Kounios
Cognitive Processes Involved in Stereotyping (Psychology), Susan T. Fiske and
Cydney H. Dupree
Language and Thought (Psychology), Susan Goldin-Meadow
Concepts and Semantic Memory (Psychology), Barbara C. Malt
Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg
Gestural Communication in Nonhuman Species (Anthropology), Simone Pika
Attention and Perception (Psychology), Ronald A. Rensink
Vocal Communication in Primates (Anthropology), Katie E. Slocombe
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
Speech Perception (Psychology), Athena Vouloumanos
Theory of Mind (Psychology), Henry Wellman



Media

Language, Perspective, and Memory