This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Explain how it is possible for people to talk about object location, or a change of object location?
The ability to understand space is vital to both animal and human survival (Landau, 2002), and the ability to talk about object location is considered a basic feature of language (Coventry & Garrod, 2004). Spatial prepositions are a relatively small class of words, but can be used in many different ways to talk about object location (Landau & Jackendoff, 1993). Projective prepositions, such as "to the left/right of", "behind/in front of", and "above/below" are one way of talking about object location, but these prepositions rely on the use of a certain perspective or frame of reference (FoR) (Coventry & Garrod, 2004). FoRs have been suggested to be a core feature of spatial language (Landau & Hoffman, 2005). The current essay will examine the three FoRs proposed by Levinson (1996), and consider how these FoRs are used in language to make it possible for people to talk about object location.
FoRs are a coordinate system allowing the use of projective prepositions to understand and talk about the location and direction of one object (the figure) in relation to another (the ground) (Coventry & Garrod, 2004; Retz-Schmidt, 1988). There are three different linguistic FoRs; intrinsic, relative and absolute (Levinson, 1996). Each of these FoRs relies on one vertical axis (above/below) and two horizontal axes (front/back and left/right). The vertical axis has been proposed to be special as it relies on an inference from gravity (Retz-Schmidt, 1988). The three FoRs normally converge on the same interpretation of vertical space, as people tend to think about objects in the upright canonical position in line with gravity (Carlson-Radvansky & Irwin, 1993; Levinson, 2003). The front-back horizontal axis has also been highlighted as special, as the asymmetric human body, allows the assignment of these spatial regions more easily than left and right (Retz-Schmidt, 1988). Franklin and Tversky (1990) found that when participants read descriptions of object locations in relation to the front, back, left, right, head (above) or feet (below) of an imagined observer, reaction times were fastest for the front/back and head/feet axes than for the left/right axis suggesting that the front/back and above/below axes may indeed be special due to their relation to gravity and body asymmetries (respectively). The fact that the right/left axis was slowest may be attributable to the fact it is "not correlated with an axis of the world, nor does it have salient asymmetries" (Franklin, Tversky & Coon, 1992, p.508). All three of these axes are, however, important as they structure each of the FoRs and allow the categorisation of space and assignment of direction in the use of projective prepositions to talk about object location.
The intrinsic FoR is object-centred, with coordinates based around the ground through which the location of the figure is described (Levinson, 1996). For example, "The ball is to the right of the chair" shows an intrinsic FoR as the ball's location is described in relation to the intrinsic right of the chair (see Figure 1). The intrinsic FoR can be considered binary as it involves talking about object locations in terms of two items: the figure and ground (Kemmerer, 2006). In order to talk about object locations, from an intrinsic FoR, the axes of the ground must be extended in order to divide space into intrinsic regions, such as "front" and "back". van der Zee and Eshuis (2003) conducted three experiments examining intrinsic axis assignment in Dutch speaking participants. Participants labelled dots on the different sides of an object to indicate the front, left, right, and back of the object. van der Zee and Eshuis found that the labels were assigned by fitting a cuboid representation around the ground, with the front/back axis being assigned to the most directionally marked axis, and the left/right axis being assigned consequently. According to van der Zee and Eshuis, directional marking is affected by ground features such as axis length, shape curvature and shape expansion (see Jackendoff, 1996a, for more information on ground features). Additionally Kemmerer (2006) highlighted that these cuboid axes are extended outwards in order to define a search area, such as "in front of" or "to the left of", in which to locate an object and therefore use the correct preposition to describe the object's location. Although these axes divide space, they do not necessarily determine direction. As van der Zee and Eshuis (2003) highlighted, front and back always occur on the same cuboid axes, but the side of the ground labelled as front or back varied. Kemmerer (2006) highlighted that the functional features of the ground object often determine this direction, and this occurs across cultures even with objects that do not appear to have a functional front or back. For example, Heine (1997) found that speakers of Chamus (Kenya) identified the front of a tree as the side that the tree leant towards. Functional features are also important in selecting which FoR to use. The presence of functional features has been found to increase the likelihood that the intrinsic FoR will be used instead of the relative or absolute FoR to talk about object location (Carlson-Radvansky & Radvansky, 1996).Additionally the locative terms used in the intrinsic FoR have been proposed to derive from body-part terms, such as "the arm/leg of a chair" (Kemmerer, 2006, p.4), and some cultures, such as Tzeltal, have been found to apply these body terms to objects, independent of the person's own viewpoint, with no apparent bodily features, such as a stone (Levinson, 1994, ). As Kemmerer (2006) highlighted this allows these intrinsic body features to be used to project a search area for the figure's location in relation to the ground.
The relative FoR is viewer-centred, representing a perspective from which both the ground and figure are being viewed (Levinson, 1996). For example, "The ball is to the left of the chair" shows a relative FoR as an independent viewpoint is used to describe the spatial location of the figure in relation to the ground (see Figure 1). To express an object's location from a relative FoR, viewers must project body axes onto the ground (Kemmerer, 2006; Levinson, 2003). The projection of body axes may involve a 180° rotation of just the front/back axis, as in English, both the front/back and left/right axes, as in Tamil (India), or no rotation of the body axes at all, as in Hausa (Nigerian) (Kemmerer, 2006) (see Figure 2). Another important aspect, unique to the relative FoR, is that of perspective sharing, which is needed in order to communicate object location to another speaker. Schober (1993) showed that when speakers in dialogue were trying to describe the same scene from different viewpoints, each participant was more likely to use a relative FoR from their own egocentric viewpoint. Feedback between speakers was important for perspective alignment as it allowed an explicit understanding of which perspective was being used to convey the object location. The importance of feedback was reiterated by Steel and Loetzch (2009) who found that robots could share perspectives easier if they were allowed to explicitly feedback the perspective that they were taking. Schober (1993) also found that both conversation partners were as likely to take the other person's perspective (an allocentric relative FoR) as each other, suggesting that participants would take their conversational partner's perspective if it helped in locating the object. Additionally, mental rotation abilities have been suggested to be an indication of a person's ability to match their perspective with their conversational partner's (Schober, 2009). Finally, use of the relative FoR also appears to be associated with living in an urban environment, although the relative FoR has been found to be used in some rural environments, and the reason for this correlation is unclear (Majid, Bowerman, Kita, Haun & Levinson, 2004).
The absolute FoR is environment-centred, based on cardinal vertical (e.g. gravity based) and horizontal (e.g. north, south, east, west) directions, or on stable environmental features, such as mountain slopes or river systems (Levinson, 1996). Kemmerer (2006) highlighted that in order to use the absolute FoR, the individual must project an angle from the ground object to the figure object and then assess object location based on appropriate terms of cardinal directions or stable environmental features. For example, "The ball is to the North of the chair" expresses the figure's location in relation to the ground through the use of a fixed cardinal direction system (see Figure 1). Both the relative and absolute FoRs can be considered tertiary systems as they involve three items: the figure, the ground, and a person- or environment-centred viewpoint (respectively). Kemmerer (2006) highlighted that the absolute system is complex and that adult's use of the absolute FoR to talk about object location requires a high level of neurocognitive effort. Levinson's (1996) description of Tzeltal revealed that speakers conveyed object location through an absolute FoR based around "uphill" and "downhill" horizontal regions as indicated by a local mountain range. The system also involves a horizontal "across" axis whereby the east or west direction is specified by environmental features. Levinson (1996) highlighted that speakers of Tzeltal can continue to use the same "uphill", "downhill" and "across" compass system when no longer in the environment in which the mountain range and usual environmental cues exist. Complexity arises as the absolute FoR is associated with the need to constantly update and monitor a mental compass relating the person and the objects around them within a system of fixed environmental or cardinal directions (Kemmerer, 2006; see Janzen & van Turennout, 2004). This requires a vast amount of cognitive effort. Nonetheless, speakers are able to talk about object locations through an absolute FoR.
There is also evidence from cognitive neuroscience that suggests a neural basis for the FoRs. The three FoRs are classified as a type of categorical spatial relation as they involve dividing space into discrete regions for talking about object locations through language (Amorapanth, Widick & Chatterjee, 2009). Neuroimaging and patient-lesion studies have consistently suggested that the left inferior parietal cortex, and in particular the supramarginal and angular gyri, are involved in categorical spatial processing (Amorapanth et al., 2009; Noordzij, Neggers, Ramsey & Postma, 2008; Tranel & Kemmerer, 2004; see Kemmerer, 2006). Additionally, Zaehle et al. (2007) found that the use of an egocentric (or relative) FoR activated regions within the posterior superior parietal lobe, and exclusively the precuneus, whilst an allocentric (object- or environment-oriented) FoR activated regions including the right inferior and superior frontal gyrus, and the right inferior and superior parietal lobe. However, Committeri et al. (2004) found that a relative FoR activated a bilateral parietal-frontal network, with greater activation in the right hemisphere, whilst the intrinsic FoR activated the ventrolateral occipital-temporal cortex, and the absolute FoR involved activation in the ventromedial occipital-temporal and retrosplenial regions. Committeri et al. suggested this absolute FoR activation reflected the matching of the current environment with stored environment representations, and so may reflect processing within a cognitive map. The bilateral activations reported in studies, whereby FoRs rely on the left hemisphere areas, as in categorical spatial processing, and right hemisphere areas, as in coordinate spatial processing (which involve continuous metrics; Amorapanth et al., 2009) may reflect the nature of FoRs underlying both these types of spatial processing. Additionally, Kemmerer (2006) proposed that the hippocampal structures may be activated with the use of an absolute FoR, as information is transferred to long-term storage as a part of the processes of updating a cognitive map, and the hippocampus has been shown to be active when long-term memories are encoded (Corkin, Amaral, González, Johnson & Hyman, 1997; Schacter, Alpert, Savage, Rauch & Albert, 1996; Scoville & Milner, 2000; Squire et al., 1992). Kemmerer (2006) also predicted that the extrastriate body area, which is involved in the visual categorisation of human body parts (Chan, Peelen & Downing, 2004), may be found to be active with intrinsic FoR as it is based around body axes. Clearly, these hypotheses need testing and much further investigation is needed into the neural areas involved in FoR use. However, the findings discussed suggest that differential activation patterns may exist for the different FoRs, suggesting a neural basis for FoRs making it possible to talk about object location.
These FoRs have been suggested to be one of the most important aspects of spatial language (Majid et al., 2004). The three FoRs can converge on the same descriptions of object locations, but can also contradict each other (see Figure 1). Therefore, in order to talk about an object location, it is necessary to select one FoR for use. Evidence suggests that, in languages where multiple FoRs are available to be used, all FoRs are initially activated before one is selected, and spatial regions are defined fastest when the FoRs coincide on an object's location (Carlson-Radvansky & Logan, 1997). Carlson-Radvansky and Jiang (1998) used a negative priming paradigm, such as that used in visual attention studies (e.g. Tipper & Cranston, 1985; see Fox, 1995, for a review), to investigate online FoR selection in English speaking participants, who can use all three FoRs (Majid et al., 2004). In Carlson-Radvansky and Jiang's (1998) study the intrinsic FoR contradicted the absolute and relative FoRs in one prime trial (Prime A), was not present in the other prime trial (Prime B), and was the target FoR in the probe trial. Participants had to judge the acceptability of the preposition "above", presented in a sentence, for the object locations presented. Carlson-Radvansky and Jiang found that when the probe trial followed Prime A, but not Prime B, it took longer to judge the acceptability of "above", suggesting that in Prime A the intrinsic FoR had been inhibited when it contradicted the other FoRs in the probe trial. Carlson-Radvansky and Jiang concluded that all FoRs are automatically initially activated when presented with spatial descriptions, but that inhibition is used to help select an appropriate FoR, by suppressing contradictory FoRs (see also Carlson, 1999; Carlson-Radvansky & Irwin, 1994). The event-related potential (ERP) technique is a useful way of measuring brain electrical activity and understanding the time course of activations associated with cognitive processes (Luck, 2005), and it has been used to examine linguistic FoRs. In an ERP study of FoR assignment, Taylor, Faust, Sitnikova, Naylor and Holcomb (2001) found that when spatial terms matched the intrinsic FoR, less semantic integration was needed, as evidenced by a decreased N400 component. The N400 has been proposed to reflect general semantic processing (Holcomb, 1993) and these findings suggest that the intrinsic FoR is prioritised in English speakers when FoR contradictions occur, as it allows more efficient linguistic interpretations of space based on less semantic information. Therefore, evidence suggests that it is possible for people to overcome contradictions between FoRs and select a particular FoR in order to talk about and understand object locations.
Given that three FoRs have been discovered, it is important to consider how these FoRs are used in different languages, and the universality of these FoRs for using projective prepositions to talk about object location. Pederson et al. (1998) used a Men and Tree game to examine FoR use across different languages. Participants had to describe the object location in one picture to the other participant, whilst making it distinct from several other similar pictures. Different combinations of the three FoRs were found to be used across several languages, but all languages used at least one FoR. Similarly, English speakers have been found to use all three FoRs (Majid et al, 2004), whilst Tzeltal speakers mainly use the absolute FoR, have some limited use of the intrinsic FoR, but never use the relative FoR (Levinson, 1996). Majid, et al. (2004) summarised FoR use in 20 different languages, and found that both the intrinsic and absolute FoR were used in each language, but that the absolute FoR was generally restricted to specialised circumstances in many languages, whilst the intrinsic FoR was consistently used in all langauges. The relative FoR was the least often used, being completely absent from some languages and only used in specialised circumstances in others. Relevant to this discussion though, the research has shown that all languages appear to use one or more of the FoRs when talking about object location. It is also important to note that there are cross-linguistic differences within FOR use. As already highlighted, rotational differences exist in different languages for use of the relative FoR, the basis of the absolute FoR differs depending on whether cardinal directions or environmental features are used, and intrinsic FoR use differs depending on whether body-axis systems or functional features are used (Kemmerer, 2006). Additionally, it should be highlighted that these FoRs can also be seen in individuals using American Sign Language, allowing them to linguistically represent object locations, although not through speech per se (Emmorey & Falgier, 1999). Therefore, FoRs appear to be present in all languages allowing people to express object locations.
As FoRs appear to be used to talk about object locations in many, and possibly all languages, it is also important to consider how these FoRs develop for use in children. Projective prepositions have been found to be learnt later in life than topological prepositions and proximity terms, and this finding appears stable across different languages (Bowerman & Choi, 2003; Johnston, 1988; Johnston & Slobin, 1979). Additionally, evidence has suggested that infants aged 3-4 months old can categorise space according to the axes present in FoRs (Behl-Chadha & Eimas, 1995; Quinn, 1994). Piaget and Inhelder (1956) found children aged 2-7 years old to be egocentric in their thinking, and the notion that spatial thinking and behaviour may be egocentric in nature has been echoed by more recent research (Wang & Spelke 2002). Importantly, around the age that children are said to be egocentric, language and particularly spatial language, begins to emerge (Barrett, 1996; Johnston, 1988). However, contrary to the notion that spatial cognition is egocentric some languages show a tendency to use absolute or intrinsic FoRs over the relative FoR. Majid et al. (2004) asked the question of whether the relative FoR is easier learnt, or dominant, in the development of the egocentric child's spatial language. Johnson (1988) highlighted that children speaking various languages, including English where the relative FoR is dominant, showed the ability to use the intrinsic FoR for "front/back" before they showed the ability to use the relative FoR. Brown and Levinson (2000) reported that Tzeltal children aged 1.5-4 years old show an understanding of the absolute FoR based around the "uphill" and "downhill" system before displaying intrinsic use of projective prepositions to talk about object locations. Brown and Levinson also found that older Tzeltal children, aged 7-8 years old, could show full use of the absolute FoR, but that Western children of the same age had not acquired the same level of use with their dominant relative FoR (Levinson, 1996). de León (1994) presented evidence that suggested that the intrinsic FoR preceded development of the relative FoR, but both de de León and Brown and Levinson (2000) suggested that the absolute FoR could develop at the same time as the intrinsic FoR. Therefore, as Majid et al. (2004) highlighted the evidence suggests that a relative FoR is not the first to develop, and that the relative FoR is not learnt more easily than the other FoRs.
It is also important to highlight that FoRs appear to be involved in non-linguistic spatial thinking, as well as in language to talk about object location (see Brown and Levinson, 1993; Landau & Hoffman, 2005; Levinson, 1996; Levinson, Kita, Haun & Rasch, 2002; Li & Gleitman, 2002). However, the debate over whether language determines thought, which is known as linguistic relativity (Kay & Kempton, 1984), or whether thought determines language is too complex to discuss here. Disentangling this relationship is especially difficult as linguistic studies generally reflect relationships between language and thought and do not allow the inference of causation. Much further research is needed to understand the relationship between spatial language and spatial thinking, but as Tversky and Lee (1998) pointed out the relationship is unlikely to be deterministic, and more likely involves thinking and language in an intertwined relationship influencing each other allowing people to think and talk about space. Therefore it seems likely that it is possible to talk about object locations using FoRs because we also think of space in terms of FoRs. It is also important to acknowledge that FoRs and projective prepositions are not only way to talk about space, and other important linguistic computations, such as those involving topological relations, such as "in" and "on", are also used to talk about object locations (Coventry, 1998; Coventry & Garrod, 2004). Further, the three FoRs discussed here have been proposed to be able to divided into much more finely-grained FoRs, and a fourth FoR for motion has been proposed (Jackendoff, 1996b). Nonetheless, the current essay has focused on the three main types of FoR, intrinsic, relative and absolute, and the evidence presented has suggested that FoRs are one means through which object location can be talked about.