SPACE IN LANGUAGE AND COGNITION Explorations in Cognitive Diversity

representation of it in terms of its volumetric properties tomatch this token to our mental inventory of such types. Although recent developments have challenged the role of the D model within a modular theory of vision, there can be little doubt that at some conceptual level such an object-centred frame of reference exists. This is further demonstrated by work on visual imagery, which seems to show that presented with a viewer-centred perspective view of a novel object, we can mentally rotate it to obtain different perspectival ‘views’ of it, for example to compare it to a prototype (Shepard andMetzler , Kosslyn , Tye : –). Thus, at some level, the visual or imagistic systems seem to employ twodistinct reference frames, viewer-centred andobject-centred. This distinction between viewer-centred and object-centred frames of reference relates rather clearly to the linguistic distinction between deictic and intrinsic perspectives discussed below: the deictic perspective is viewer-centred, while the intrinsic perspective seems to use (at least partially) the same axial extraction that would be needed to compute the volumetric properties of objects for visual recognition (see Landau and Jackendoff , Jackendoff , also Levinson ). This parallel will be further reinforced by the reformation of the linguistic distinctions suggested in the section below. This brings us to the distinction between orientation-bound vs. orientation-free frames of reference. The visual imagery and mental rotation literature might be thought to have little to say about frames of reference. After all, visual imagery would seem to be necessarily at most .D and thus necessarily in a viewer-centred frame of reference (even if mental rotations indicate access to a D structural description). But recently there have been attempts to understand the relation between two kinds of shape recognition: the process where shapes can be recognized without regard to orientation (thus with no response-curve latency related to angular displacement from a familiar related stimulus), and another process where shapes are recognized by apparent analogue rotation to the familiar related stimulus. The Shepard andMetzler paradigm suggested that only where handedness information is present (as where enantiomorphs have to be discriminated) would mental rotation be involved, which implicitly amounts to some distinction between object-centred and viewer-centred frames of reference: discrimination of enantiomorphs depends on an orientation-bound perspective, while the recognition of simpler shapes may be orientation-free. But some recent controversies seem to show that things are not as simple as this (Tarr and Pinker , Cohen and Kubovy ). Just and Carpenter () argue ‘Frames of reference’ across modalities  that rotation tasks in fact can be solved using four different strategies, some orientation-bound and some orientation-free. Similarly, Takano () insists that only orientation-bound forms should require mental rotation for recognition. However, Cohen and Kubovy () claim that all this makes the wrong predictions since handedness-identification can be achieved without the mental-rotation latency curves in special cases. In fact, I believe that, despite these recent controversies, the original assumption – that only objects lacking handedness can be recognized without mental rotation – must be basically correct for logical reasons that have been clear for centuries. In any case, it is clear from this literature that the study of visual recognition and mental rotation utilizes distinctions in frames of reference that can be put into correspondence with those that emerge from, for example, the study of language: absolute and relative frames of reference in language (to be firmed up below) are both orientation-bound, while the intrinsic frame is orientation-free (Danziger ). Linguists have long distinguished ‘deictic’ vs. ‘intrinsic’ frames of reference, because of the rather obvious ambiguities of a sentence likeThe boy is in front of the house – the boy can be at the house’s front, or the boy can be between the speaker and any side of the house (see, e.g., Leech : , Fillmore , Clark ). It has also been known for a while that linguistic acquisition of these two readings of terms like in front, behind, to the side of is in the reverse direction from the developmental sequence ‘egocentric’ to ‘allocentric’ (Pick ): ‘intrinsic’ notions come resolutely earlier than ‘deictic’ ones ( Johnston and Slobin ). Sometimes a third term ‘extrinsic’ is opposed, to denote, for example, the contribution of gravity to the interpretation of words like above or on. But unfortunately the crucial term ‘deictic’ breeds confusions. In fact there have been at least three distinct interpretations of the ‘deictic’ vs. ‘intrinsic’ contrast, as listed in Table .: (a) speaker-centric vs. non-speaker-centric (Levelt ), (b) centred on any of the speech participants vs. not so centred (Levinson ), (c) ternary vs. binary spatial relations (implicit in Levelt , , the view to be adopted here). These issues will be taken up in the section below, where we will turn to ask what distinctions in frames of reference are grammaticalized or lexicalized in different languages. Let us turn now to the various distinctions suggested in the psychology of language.Miller and Johnson-Laird (), drawing on earlier linguistic work, explored the opposition between ‘deictic’ and ‘intrinsic’ interpretations of such utterances asThe cat is in front of the truck; and the logical  Frames of reference properties of these two frames of reference, and their interaction, have been further clarified by Levelt (, , ). Carlson-Radvansky and Irwin (: ) summarize the general assumption in psycholinguistics as follows: Three distinct classes of reference frames exist for representing the spatial relationships among objects in the world . . .: viewer-centred frames, objectcentred frames, and environment centred frames of reference. In a viewer-centred frame, objects are represented in a retinocentric, head-centric or body-centric coordinate system based on the perceiver’s perspective of the world. In an object-centred frame, objects are coded with respect to their intrinsic axes. In an environment-centred frame, objects are represented with respect to salient features of the environment, such as gravity or prominent visual landmarks. In order to talk about space, vertical and horizontal coordinate axes must be oriented with respect to one of these reference frames so that linguistic spatial terms such as ‘above’ and ‘to the left of’ can be assigned. These three frames of reference, renamed the relative, intrinsic and absolute respectively, are essentially those that will be adopted as a framework for the analyses in this book. But notice that on this particular formulation frames of reference inhere in spatial perception and cognition rather than in language: above may simply be semantically general over the different frames of reference, not ambiguous (Carlson-Radvansky and Irwin (: ). Thus the corresponding three-way distinctions between, for example, the ‘deictic’, ‘intrinsic’ and ‘extrinsic’ are merely alternative labels for the linguistic interpretations corresponding, respectively, to viewer-centred, object-centred and environment-centred frames of reference. There are other oppositions that psycholinguists employ, although in most cases theymap onto the same triadic distinction. One particular set of distinctions, between different kinds of survey or route description, is worth unravelling because it has caused no little confusion. Levelt (: ff.) points out that when a subject describes a complex visual pattern the linearization of speech requires that we ‘chunk’ the pattern into units that can be described in a linear sequence. Typically, we seem to represent D or D configurations through a small window, as it were, traversing the array; that is, the description of complex static arrays is converted into a description of motion through units or ‘chunks’ of the array. Levelt () has examined the description of D arrays and found two strategies: a gaze tour perspective – effectively the adoption of a fixed viewpoint where one’s gaze travels over the path (a ‘deictic’ or viewer-centred perspective) – and a body or ‘driving’ tour – effectively ‘Frames of reference’ across modalities  an intrinsic perspective, where a pathway is found through the array, and the imagined tour of oneself along the path is used to assign ‘front’, ‘left’ etc. from any one point (or location of the window in describing time). Since both perspectives can be thought of as egocentric, Tversky () opts to call Levelt’s intrinsic perspective ‘deictic frame of reference’ or ‘route description’ and his ‘deictic’ perspective she labels ‘survey perspective’. Thus Tversky’s ‘deictic’ is Levelt’s ‘intrinsic’ or non-deictic perspective! This confusion is, I believe, not merely terminological, but results from the failure in the literature to distinguish coordinate systems from their origins or centres, as discussed in the next section. There is a final issue of some importance. In psycholinguistic discussions about frames of reference, there seems to be some unclarity, or sometimes overt disagreement, about at which level – perceptual, conceptual or linguistic – such frames of reference apply. Thus CarlsonRadvansky and Irwin (, quoted above) make the assumption that a frame of reference must be adopted within some spatial representation system, as a precondition for co-ordinating perception and language, whereas Levelt (, ) has argued that a frame of reference is freely chosen in the very process of mapping from perception or spatial representation to language. On the latter conception, frames of reference in language are peculiar to the nature of the linear, propositional representation system that underlies linguistic semantics: they are different ways of conceiving the same percept in order to talk about it. The view that frames of reference in linguistic descriptions are adopted in the mapping from spatial representation or perception to language seems to suggest that the perceptions or spatial representations themselves are frame-of-reference-free. But this of course is not the case: there has to be some coordinate system involved in any spatial representation of any intricacy, whether at a peripheral, or sensory, level or at a central, or conceptual, level. What Levelt’s () results and Friederici and Levelt () seem to establish is that frames of reference at the perceptual or spatial conceptual level do not necessarily determine frames of reference at the linguistic level. This is exactly what one might expect: language is flexible and it is an instrument of communication – thus it naturally allows us, amongst other things, to take the other guy’s perspective. Further, the ability to cast a description in one frame or another implies an underlying conceptual ability to handle multiple frames, and within strict limits to convert between them (a matter to which we will return). In any case, we need to distinguish in discussions of frames of reference between at least three levels, perceptual, conceptual and linguistic, and  Frames of reference we need to consider the possibility that we may utilize distinct frames of reference at each level (although I shall later argue that they tend to be brought into congruence). There is much further pertinent literature in all the branches of psychology and brain science, but it should already be clear that there are many different classifications and different construals of the same terms, not to mention many unclarities and many deep confusions in all of this. However, despite this forest of distinctions with obscuring undergrowth, there are some obvious common bases to the distinctions we have reviewed. It is clear, for example, that, on the appropriate construals, ‘egocentric’ corresponds to ‘viewer-centred’ and ‘.D’ sketch to ‘deictic’ frame, while ‘intrinsic’ maps onto ‘object-centred’ or ‘D model’ frames of reference, while ‘absolute’ is related to ‘environmentcentred’, and so on. We should build on these commonalities, especially as in this book we will be concerned with how frames of reference in language may reflect, or induce, frames of reference in other kinds of mental representation. However, before proposing an alignment of these distinctions across the board, it is essential to deal with linguistic frames of reference, which present a troubling flexibility that has led to various confusions.  .                         Cursory inspection of the linguistic literaturewill give the impression that the linguists have their house in order. They talk happily of topological vs. projective spatial relators (e.g. as pronouns like in vs. behind), deictic vs. intrinsic usages of projective prepositions, and so on (see, e.g., Bierwisch , Lyons , Herskovits , Vandeloise , and psycholinguists Clark , Miller and Johnson-Laird ). But the truth is less comforting. The analysis of spatial terms in familiar European languages remains deeply confused, and those in other languages almost entirely unexplored. Thus the various alleged universals should be taken with a great pinch of salt – indeed many of them can be directly jettisoned. Onemajor upset is the recent finding (exemplified in Chapter  below) that many languages use an ‘absolute’ frame of reference (involving fixed bearings like ‘West’) where European languages would use a ‘relative’ or viewpoint-centred one (using notions like ‘left’). Another is that some languages, like many Australian ones, use such frames of reference to replace so-called ‘topological’ notions like ‘in’, ‘on’ or ‘under’. Linguistic frames of reference in cross-linguistic perspective  A third is that expressions for familiar spatial notions like ‘left’ and ‘right’, and even sometimes ‘front’ and ‘back’, are missing from many, perhaps a third of all languages. Confident predictions and assumptions can be found in the literature that no such languages occur (see, e.g., Clark , Miller and Johnson-Laird , Lyons : ). These developments call for some preliminary typology of the frames of reference that are systematically distinguished in the grammar or lexicon of different languages (with the caveat that we still know only a little about only a few of them). In particular, we will focus on what we seem to need in the way of coordinate systems and associated reference points to set up a cross-linguistic typology of the relevant frames of reference. In what follows I shall confine myself to linguistic descriptions of static arrays, and concentrate just on the central frames of reference, leaving a broader review of linguistic spatial systems for Chapter . Moreover, I will focus on distinctions on the horizontal plane. This is not whimsy: perceptual cues for the vertical may not always coincide, but they overwhelmingly converge, giving us a good universal solution to one axis. But the two horizontal coordinates are up for grabs: there simply is no corresponding force like gravity on the horizontal. Consequently there is no simple solution to the description of horizontal spatial patterns, and languages diverge widely in their solutions to this basic problem: how to specify angles or directions on the horizontal. Essentially, three main frames of reference emerge from this new data as solutions to the problem of description of horizontal spatial directions. They are appropriately named ‘intrinsic’, ‘relative’ and ‘absolute’, even though these terms may have a somewhat different interpretation from some of the construals reviewed in the section above. Indeed the linguistic frames of reference potentially cross-cut many of the distinctions in the philosophical, neurophysiological, linguistic and psychological literatures for one very good reason. The reason is that linguistic frames of reference cannot be defined by reference to the nature of the origin of the coordinate system (in contrast to, e.g., ‘egocentric’ vs. ‘allocentric’). It will follow that the traditional distinction ‘deictic’ vs. ‘intrinsic’ collapses – these are not opposed terms. All this requires some explanation. We may start by noting the difficulties we get into by trying to make the distinction between ‘intrinsic’ and ‘deictic’. Levelt (: –) organizes and summarizes the standard assumptions in a useful way that illustrates the problem: we can cross-classify linguistic uses according to (a) whether they presume that the coordinates are centred on the speaker or not, (b) whether the relatum or ground is the speaker or not. Suppose  Frames of reference then we call the usage ‘deictic’ just in case the coordinates are centred on, or have their origin in, the speaker, ‘intrinsic’ otherwise. This yields the following classification of examples: () The ball is in front of me Coordinates: ‘Deictic’ Origin: Speaker Relatum (Ground): Speaker () The ball is in front of the tree Coordinates: ‘Deictic’ Origin: Speaker Relatum (Ground): Tree () The ball is in front of the chair (at the chair’s front) Coordinates: ‘Intrinsic’ Origin: Not the speaker, but the chair Relatum (Ground): Chair Clearly it is the locus of the origin of the coordinates that is relevant to the traditional opposition ‘intrinsic’ vs. ‘deictic’, otherwise we would group () and () as both sharing a non-deictic relatum. The problem comes when we pursue this classification further: () The ball is in front of you Coordinates: ‘Intrinsic’ Origin: Not the speaker, but the addressee Relatum: Addressee () The ball is to the right of the lamp, from your point of view Coordinates: ‘Intrinsic’ Origin: Not the speaker, but the addressee Relatum: Lamp Here the distinction between ‘intrinsic’ vs. ‘deictic’ is self-evidently not the right classification, as far as frames of reference are concerned. Clearly, () and () belong together: the interpretation of the expressions is the same, with the same coordinate systems, there are just different origins, speaker and addressee respectively (moreover, in a normal construal of ‘deictic’, inclusive of first and second persons, both are ‘deictic’ origins). Similarly, in another natural grouping, () and () should be classed together: they have the same conceptual structure, with a viewpoint (acting as the origin of the coordinate system), a relatum distinct from the viewpoint, and a referent – again the origin (or viewpoint here) alternates over speaker or addressee. Linguistic frames of reference in cross-linguistic perspective  Wemight be tempted to just alter the designations, and label (), (), () and () all ‘deictic’ as opposed to () ‘intrinsic’. But this would be a further confusion. First, it would conflate the distinct conceptual structures of our groupings () and () vs. () and (). Secondly, the conceptual structure of the coordinate systems in () and () is in fact shared with (). How? Consider: The ball is in front of the chair presumes (on the relevant reading) an intrinsic front, and uses that facet to define a search domain for the ball; but just the same holds for The ball is in front of me/you. Thus the logical structure of (), () and () is the same: the notion ‘in front of ’ is here a binary spatial relation, with arguments constituted by the figure (or referent) and the ground (or relatum), where the projected angle is found by reference to an intrinsic or inherent facet of the ground object. In contrast, () and () have a different logical structure: ‘in front of’ and ‘to the right of ’ are here ternary relations, presuming a viewpointV (the origin of the coordinate system), a figure andground, all distinct. In fact, these two kinds of spatial relation have quite different logical properties, as demonstrated by Levelt (, ), but only when distinguished and grouped in this way (more in a moment). Let us dub the binary relations ‘intrinsic’, but the ternary relations ‘relative’ (because the descriptions are always relative to a viewpoint, in contradistinction to ‘absolute’ and ‘intrinsic’ descriptions). To summarize then, the proposed classification (retaining the earlier numbering of examples) is: () The ball is in front of me Coordinates: Intrinsic Origin: Speaker