Elsevier

Cognitive Psychology

Volume 56, Issue 2, March 2008, Pages 142-163
Cognitive Psychology

Is language necessary for human spatial reorientation? Reconsidering evidence from dual task paradigms

https://doi.org/10.1016/j.cogpsych.2007.06.002Get rights and content

Abstract

Being able to reorient to the spatial environment after disorientation is a basic adaptive challenge. There is clear evidence that reorientation uses geometric information about the shape of the surrounding space. However, there has been controversy concerning whether use of geometry is a modular function, and whether use of features is dependent on human language. A key argument for the role of language comes from shadowing findings where adults engaged in a linguistic task during reorientation ignored a colored wall feature and only used geometric information to reorient [Hermer-Vazquez, L., Spelke, E., & Katsnelson, A. (1999). Sources of flexibility in human cognition: Dual task studies of space and language. Cognitive Psychology, 39, 3–36]. We report three studies showing: (a) that the results of Hermer–Vazques et al. [Hermer-Vazquez, L., Spelke, E., & Katsnelson, A. (1999). Sources of flexibility in human cognition: Dual task studies of space and language. Cognitive Psychology, 39, 3–36] are obtained in incidental learning but not with explicit instructions, (b) that a spatial task impedes use of features at least as much as a verbal shadowing task, and (c) that neither secondary task impedes use of features in a room larger than that used by Hermer–Vazquez et al. These results suggest that language is not necessary for successful use of features in reorientation. In fact, whether or not there is an encapsulated geometric module is currently unsettled. The current findings support an alternative to modularity; the adaptive combination view hypothesizes that geometric and featural information are utilized in varying degrees, dependent upon the certainty and variance with which the two kinds of information are encoded, along with their salience and perceived usefulness.

Introduction

In order to survive, humans and all other mobile animals must be able to locate desirable objects in their surrounding environment, and avoid areas of danger. Spatial orientation and navigation abilities seem to rely on two major coding systems: a self-based “dead reckoning” system that tracks the location of the self in relation to the environment as the self moves, and an environment-based allocentric system, in which location is coded in terms of the surroundings (Gallistel, 1990, Newcombe, 2002, Sholl, 1992). These systems are complementary, with the dead reckoning system proving vital when landmarks are lacking, as in the dark or on the open ocean. However, the allocentric system is required for avoiding the inevitable drift that occurs in dead reckoning, created by the concatenation of small errors in judging distance and the angles of turns. This environment-centered spatial orientation system is further distinguished by its utilization of two types of spatial information: geometric and nongeometric. The shape of a landmark is typically regarded as geometric information, while all other characteristics of the landmark, such as color, texture, and size, are regarded as nongeometric or featural information.

A unique situation occurs when a mobile creature has lost the ability to use the dead reckoning system, due, for example, to rapid and erratic turns or (in humans) passive underground travel as by subway. In this case, environmental information is clearly required to re-establish orientation. However, interestingly, not all environmental information appears to be created equal for disoriented animals. Cheng (1986) examined how rats searched for previously located food hidden in one corner of a rectangular enclosure after a disorientation procedure of removing and replacing the rat in a different position. He found that rats searched the geometrically equivalent corners in an unfeatured enclosure, that is, they systematically chose the correct corner and the corner diagonally opposite in a featureless rectangular environment, showing that they encoded the geometric properties of the space including metric and sense information (length of short versus long walls and left-right relationships, respectively). Further, Cheng found that when he added features that would allow the rats to distinguish between the two geometrically equivalent corners, such as walls differing in shade or patterned corners, the rats continued to divide their search evenly between the two geometrically equivalent corners when the correct location changed across trials. However, when the correct location remained constant, the rats did learn to use the features but only in the case of a direct cue at the correct location. Once the distinguishing feature at the target location was removed, the rats did not use the remaining cues in surrounding corners, reverting to a more geometric reorientation strategy. That is, geometric information was seemingly always used to reorient, whereas featural information was secondary and had limited use among disoriented rats.

From these data, the proposition of a geometric module designated for spatial orientation emerged (Cheng, 1986, Cheng and Gallistel, 1984). Such a module would serve to process the geometric properties of an organism’s surroundings, determined by metric and sense information, in order to guide spatial navigation. The module is encapsulated, in Fodor’s (1983) sense, in that geometric information is the only source utilized for spatial representations and subsequent navigation. Even though nongeometric information may allow one to correctly identify a specific location, such as using a colored wall to distinguish between two geometrically correct corners in a rectangular enclosure, this information is not considered during reorientation within a modular view. Gallistel (1990) argued that the existence of such an encapsulated module might be adaptive because geometry is often preserved when features change, e.g., when the shape of a river bank is maintained despite changes in the muddiness of the water.

Hermer and Spelke, 1994, Hermer and Spelke, 1996 examined the possible existence of a geometric module in humans. They found that human children, when asked to search for a toy that they had watched the experimenter hide in one corner of a rectangular room, behaved much like Cheng’s rats. After being spun in circles with their eyes closed to become disoriented, children as young as 18 months searched for a hidden toy in the two geometrically equivalent corners of an unfeatured room, suggesting a remarkable sensitivity to relative length and sense of the walls. However, children until the age of 6 years did not use nongeometric information, such as colored walls, even when given the opportunity.

One explanation for this shift in use of features between the age of 5 and 6 years is that children then become able to use the linguistic terms “right” and “left”. In support of this idea, Hermer-Vazquez, Moffet, and Munkholm (2001) found a correlation between use of these terms and use of features in the reorientation task. However, correlational evidence is inherently weak, being subject to third variable problems. More striking evidence for the role of language in the ability to overcome an inherently modular cognitive architecture was reported by Hermer-Vazquez, Spelke, and Katsnelson (1999). They found that adults simultaneously performing a verbal shadowing task while also searching for objects following disorientation behave like children and rats, failing to use a colored wall to constrain searches in a rectangular room. The participants did, however, use the colored wall while simultaneously performing a nonverbal rhythm-clapping task, suggesting that it was language rather than simple cognitive overload that impeded the ability to use featural information. These results have seemed to provide strong support for the conclusion that language is necessary for allowing adults to overcome the encapsulation of the geometric module.

More recently, however, there have been questions concerning the proposed modular architecture and about the necessity of language for supporting spatial reorientation across various species. First, studies with non-human (and hence non-linguistic) animals have cast doubt on these two related ideas, by showing that features are sometimes used as well as geometry in the reorientation task (Chiandetti et al., 2007, Sovrano and Vallortigara, 2006, Vallortigara et al., 1990 for chickens; Kelly, Spetch, & Heth, 1998 for pigeons; Gouteux, Thinus-Blanc, & Vauclair, 2001 for monkeys; Sovrano et al., 2002, Sovrano et al., 2003, Sovrano et al., 2005, Sovrano et al., 2007, for fish). However, while these findings from animal research might seem to provide strong evidence against the geometric module hypothesis, particularly against the idea that language is necessary to integrate geometric and nongeometric information, Sovrano et al. (2003) point out that though some nonverbal animals are successful in completing these spatial tasks, this does not necessarily mean that they use the same mechanisms to perform the tasks that humans would use. Similarly, Hermer-Vazquez et al. (2001) have suggested that these results from non-human studies might be a reflection of the extensive training typically found when working with animals. Thus, direct evidence from humans is necessary to support this argument. Not only is evidence from humans warranted in order to understand how and under what circumstances different spatial cues are utilized, such evidence would verify if indeed human and non-human reorientation systems are homologous.

In a series of studies, Learmonth, Newcombe, and Huttenlocher (2001) found that children as young as 18 months do use featural landmarks in addition to the shape of the room to successfully reorient and find a hidden toy. Learmonth et al. replicated Hermer and Spelke, 1994, Hermer and Spelke, 1996 finding that disoriented children use geometric information to search at the two geometrically equivalent corners of a featureless rectangular room, using an area four times that of the original Hermer and Spelke studies. However, when features were added to the larger room, such as a door to the room on one long wall and a recessed bookcase on the other long wall, children between 18 and 24 months used these landmarks, in addition to the shape of the room, in their search for the toy. Further experiments established that children could successfully reorient using a single landmark (including a colored wall) as well, suggesting that human children can integrate geometric and nongeometric information to successfully complete these spatial tasks prior to acquiring spatial language.

A key factor leading to the difference in results was the size of the room (Learmonth, Nadel, & Newcombe, 2002). Hermer and Spelke had used a very small room, but more distal features are more valuable spatial cues and hence might be more likely to be used (Nadel & Hupbach, 2006; see also Wang & Spelke, 2002). Further evidence of the importance of distal versus proximal landmarks on influence of feature use is found in recent work with non-human animals. Fish make relatively more geometric errors when reorienting in a small tank and rely more on features when reorienting in a large tank (Sovrano et al., 2007). Similarly, when faced with conflicting geometric and featural cues from a learned spatial representation, chicks (Chiandetti et al., 2007, Sovrano and Vallortigara, 2006, Vallortigara et al., 2005) and fish (Sovrano et al., 2007) rely on geometry to reorient in a small enclosure but use features to a greater extent in a larger environment. There is even some neurobiological evidence to support the importance of proximity when utilizing features, in that the head-direction cells of rats seem to depend on information from distal rather than proximal cues (Zugaro et al., 2004).

Although it is interesting and important for the understanding of spatial development to answer the question of why young children use features in addition to geometry in the larger room but not the small room, these results still provide evidence that children and even non-human animals are not limited to geometric information when reorienting in a rectangular enclosure. The finding itself, that spatial reorientation occurs in these various groups, is of vital importance to the debate on the very existence of a geometric module. Additionally, the data from very young children and non-human animals raise serious concerns for the suggestion that acquiring spatial language is essential for the ability to combine featural and geometric information.

Nevertheless, the striking data of Hermer-Vazquez et al. (1999) require explanation. Why is it that human adults required to do a verbal shadowing task (but not a control task that seemed equally attention-demanding) fail to use a feature as large as a colored wall to guide search? We suggest that there are two possible ways to explain these results. First, in the Hermer-Vazquez et al. study (1999), adults were simply informed prior to the disorientation procedure that, “you will see something happening that you should try to notice,” and that they would be asked about what they saw. Following these vague instructions, and with no practice trials, the search task in a rectangular room with a blue wall occurred, followed by search without shadowing in that room, and by search in an all-white room. Order was not counterbalanced. It is possible that, if given a clearer idea of the demands of the reorientation task, adult participants could search the correct corner at greater than chance levels even while engaged in verbal shadowing. We examine this idea in Experiment 1.

Second, the verbal shadowing task used by Hermer-Vazquez et al. (1999) might disrupt the ability to use featural landmarks not (or not only) by interfering with a linguistic encoding process within a geometric module, but by interfering with a non-linguistic, nonmodular spatial encoding system, whereby geometric and nongeometric information is weighted depending on the certainty and variance with which the two kinds of information are encoded and then utilized accordingly. The nonverbal rhythm-clapping task used by Hermer-Vazquez et al. (1999) might be ill-suited to examine this possibility because it involves primarily cerebellar regions of the brain (Woodruf-Pak, Papka, & Ivry, 1996) and would not be expected to engage spatial coding systems. A nonverbal spatial task might, however, interfere with the integration of geometric and featural information in the reorientation task (Newcombe, 2005). We look at this issue in Experiment 2.

Experiment 3 in this paper examines whether linguistic and/or spatial tasks affect the use of features to reorient in a larger room than that used in the Hermer-Vazquez studies. One potential way to explain the fact that successful use of features begins earlier in larger rooms would be to postulate that the mechanisms needed to combine features and geometry differ as a function of the size of the space. In particular, perhaps language is helpful (if not absolutely necessary) in the smaller room, but not essential in the larger one. We consider these possibilities in Experiment 3.

Section snippets

Experiment 1

One condition of this experiment was a straightforward replication of Hermer-Vazquez et al. (1999) basic paradigm (their Experiment 1) in which participants received vague task instructions, no practice trial and a fixed order of conditions beginning with shadowing in the presence of the colored wall. In a second condition, participants were given additional instructions about the nature of the search task, performed a practice trial prior to performing the reorientation task, and then

Experiment 2

In order to examine the effects of interference tasks on incidental learning among adults, we again replicated the reorientation task used by Hermer-Vazquez et al. (1999). However, we substituted a different nonverbal shadowing task. The nonverbal rhythm-shadowing condition they used may be an inappropriate control for the verbal shadowing task because keeping a rhythm involves primarily cerebellar regions of the brain, and hence would not be expected to be centrally involved with spatial

Experiment 3

Experiments 1 and 2 suggest that concurrent tasks, whether linguistic or spatial in nature, disrupt encoding of features in an incidental condition. These experiments were conducted in a small room, only 4 by 6 feet, in order to replicate the conditions of the previous studies. However, Learmonth et al., 2001, Learmonth et al., 2002 have found that young children are more likely to use features in a larger room. To examine whether integration of geometric and nongeometric information has

References (49)

  • V.A. Sovrano et al.

    Modularity and spatial reorientation in a simple mind: Encoding of geometric and nongeometric properties of a spatial environment by fish

    Cognition

    (2002)
  • R. Wang et al.

    Human spatial representation: Insights from animals

    Trends in Cognitive Sciences

    (2002)
  • M.S. Banks

    The benefits and costs of combining information between and within senses

  • S. Benhamou et al.

    Landmark use by navigating rats (Rattus norvegicus) contrasting geometric and featural information

    Journal of Comparative Psychology

    (1998)
  • L. Brooks

    Spatial and verbal components of the act of recall

    Canadian Journal of Psychology

    (1968)
  • A.A. Brown et al.

    Growing in circles: Rearing environment alters spatial navigation in fish

    Psychological Science

    (2007)
  • K. Cheng et al.

    Testing the geometric power of a spatial representation

  • K. Cheng et al.

    Is there a geometric module for spatial orientation? Squaring theory and evidence

    Psychonomic Bulletin & Review

    (2005)
  • C. Chiandetti et al.

    Spatial reorientation: The effects of space size on the encoding of landmark and geometry information

    Animal Cognition

    (2007)
  • M. Ernst et al.

    Humans integrate visual and haptic information in a statistically optimal way

    Nature

    (2002)
  • J. Fodor

    Modularity of mind: An essay on faculty psychology

    (1983)
  • C.R. Gallistel

    The organization of learning

    (1990)
  • S. Gouteux et al.

    Rhesus monkeys use geometric and nongeometric information during a reorientation task

    Experimental Psychology: General

    (2001)
  • E.R. Gray et al.

    Spatial encoding in mountain chickadees: Features overshadow geometry

    Biology Letters

    (2005)
  • Cited by (76)

    • Separation of geometric and featural information in children's spatial representation: Evidence from a model selection task

      2022, Journal of Experimental Child Psychology
      Citation Excerpt :

      Second, previous studies on reorientation (e.g., Hermer-Vazquez, Spelke, & Katsnelson, 1999) suggested that humans’ ability to combine diverse information into unitary representations depends on natural language. Although there are some debates on this speculation (e.g., Ratliff & Newcombe, 2008), it is worthwhile to examine the relation between children’s performance in our model selection task and their spatial language. Finally, this study provides evidence about children’s representation of the surrounding environment, which is a basis for many spatial tasks and not only for reorientation.

    • The relations among navigation, object analysis, and magnitude perception in children: Evidence for a network of Euclidean geometry

      2020, Cognitive Development
      Citation Excerpt :

      Nevertheless, other research on spatial language (Hermer-Vazquez et al., 2001; Shusterman et al., 2011) and mapping (Dillon & Spelke, 2018; Shusterman et al., 2008) suggest alternative mechanisms by which children may use geometric cues flexibly for different purposes. Children’s knowledge of spatial-relational words predicted their performance on a navigation task (Hermer-Vazquez et al., 2001) and adults' reorientation performance was disrupted by linguistic interference (Hermer-Vazquez et al., 1999; but, see Ratliff & Newcombe, 2008a), suggesting that language may contribute to flexible geometry use beginning at 5–6 years and continuing into adulthood. Other studies on children’s mapping ability suggest a more protracted developmental trajectory in children’s use of symbols to support flexible geometry use (Dillon & Spelke, 2018).

    View all citing articles on Scopus

    Portions of this research were presented at the Psychonomic Society 2004 and at the Cognitive Science Society 2005. Funding was provided by the National Science Foundation (BCS0414302).

    View full text