File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/metho/94/p94-1042_metho.xml
Size: 24,668 bytes
Last Modified: 2025-10-06 14:13:54
<?xml version="1.0" standalone="yes"?> <Paper uid="P94-1042"> <Title>A COMPUTATIONAL VIEW OF THE COGNITIVE SEMANTICS OF SPATIAL PREPOSITIONS*</Title> <Section position="4" start_page="0" end_page="0" type="metho"> <SectionTitle> THE SEMANTICS OF PROJECTIVE PREPOSITIONS </SectionTitle> <Paragraph position="0"> In this section we characterize the components of the spatial meaning of projective prepositions that have motivated our interest in cognitive linguistic approaches. Throughout, the decoding problem, that is, generating adequate meanings for a locative expression in a particular situation, is our benchmark for representational adequacy.</Paragraph> <Paragraph position="1"> The spatial meaning Of a projective prepositional predication (eg. &quot;the chair is in front of the desk&quot;) can include: a constraint on the proximity of the located (LO) (eg. &quot;the chair&quot;) and reference objects (RO) (eg. &quot;the desk&quot;); a directional constraint on the LO relative to the RO; and a relative orientation between the speaker, LO and RO. Constraints are of an intrinsically fuzzy nature such that different relative positions and orientations of the speaker, RO and LO satisfy the predication to different degrees, and combinations of constraints on the RO and LO originating from different predications must be readily accommodated. null</Paragraph> </Section> <Section position="5" start_page="0" end_page="0" type="metho"> <SectionTitle> PROXIMITY CONSTRAINTS </SectionTitle> <Paragraph position="0"> Projective prepositions necessarily place a constraint on the proximity of the located object and the reference object. Predications such as &quot;the chair is in front of the desk&quot; constrain the &quot;desk&quot; and &quot;chair&quot;, to some degree, to be proximal to each other. Conversely projective prepositions such as &quot;away from&quot; predicate a distal relationship between the located and reference object. The degree of the proximity expressed in any projective prepositional predication varies accord- null INTRINSIC In the intrinsic case the reference frame is centered at the R0 and adopts the intrinsic orientations of the RO. Thus a LO is deemed to be &quot;in front of&quot; the RO under.an intrinsic reading if it is located in the direction defined by the vector that is the half-plane of the front of the R0.</Paragraph> <Paragraph position="1"> In figure 1 stool number I is intrinsically &quot;in front of the desk&quot;.</Paragraph> <Paragraph position="2"> DEICTIC The reference frame for a deictic interpretation is centered at the speaker and adopts the speaker's orientation; deictic readings can be invoked explicitly with qualifications such as &quot;from where we are standing&quot;; when the RO has no intrinsic or extrinsic sideness relating to the preposition used; or when intrinsic or extrinsic interpretations are ruled out on other grounds (eg.</Paragraph> <Paragraph position="3"> the impossibility of spatially arranging the objects as required by the interpretation). In figure 1 stool number 2 is deictically &quot;in front of the desk&quot;. &quot;in front off' ing to a number of considerations including: the spatial context (the spatial extent and content of the scene described); and the absolute and relative sizes of the LO and RO (eg. a car that is &quot;left of&quot; a lorry is typically less proximal than an apple and orange similarly described).</Paragraph> </Section> <Section position="6" start_page="0" end_page="0" type="metho"> <SectionTitle> DIRECTIONAL CONSTRAINTS </SectionTitle> <Paragraph position="0"> In addition to the constraint on the proximity of the LO and RO, projective prepositions place a constraint on the position of the LO relative to a particular side of the RO. In the case of the intrinsic interpretation (see section ) of a predication such as &quot;the stool is in front of the desk&quot;, the &quot;stool&quot; is located in some region of the space defined by the half-plane that is the intrinsic front of the &quot;desk&quot;. Intuitively, the closer the &quot;stool&quot; is to the region of space defined by the projection of the desk's dimensions into this space, the more the spatial arrangement conforms to the prototypical interpretation of the predication.</Paragraph> </Section> <Section position="7" start_page="0" end_page="0" type="metho"> <SectionTitle> REFERENCE FRAMES </SectionTitle> <Paragraph position="0"> Intrinsic, deictic and extrinsic interpretations of projective prepositions differ according to the reference frame with respect to which the directional constraint is characterized \[Retz-Schmidt1988\].</Paragraph> <Paragraph position="1"> Figure 1 is an example of a scene that might give rise to predications which invoke each of these reference frames.</Paragraph> <Paragraph position="2"> EXTRINSIC Extrinsic readings can occur when the RO has no intrinsic sides relating to the locative preposition (eg. for objects such as trees) but is in close proximity to another object that is strongly sided (eg. such as a house); in which case the reference frame capturing the intrinsic orientations of the stronger sided object can be adopted by the RO. Referring to figure 1 the chair is extrinsically &quot;in front of stool number 3&quot;; here the stool has inherited an extrinsic front from the right wall.</Paragraph> </Section> <Section position="8" start_page="0" end_page="0" type="metho"> <SectionTitle> INTERACTING CONSTRAINTS </SectionTitle> <Paragraph position="0"> Typically an object is located with respect to more than one RO by the means of multiple spatial predications. This places a requirement of on the meaning representation of spatial predications that they must capable of being easily combined, to give rise to a cumulative meaning.</Paragraph> </Section> <Section position="9" start_page="0" end_page="304" type="metho"> <SectionTitle> COGNITIVE GRAMMAR AND LINGUISTIC MEANING </SectionTitle> <Paragraph position="0"> Cognitive granlmar is comprised of five basic claims as to the composition of linguistic meaning, following \[Langacker1988b\] these are: 1. Meaning reduces to conceptualization. 2. Polysemy is the norm and can be adequately accommodated by representing the meaning a lexical item as a network of senses related by categorizing relationships of schematicity or extension. null 3. Semantic structures are characterized relative to cognitive domains. Domains are hierarchically organized in terms of conceptual complexity, where the characterization of a concept at one level can draw on lower level concepts. While there need not necessarily be any conceptual primitives, the lowest level domains are termed basic domains and include our experience of time, space, color etc.</Paragraph> <Paragraph position="1"> 4. A semantic structure derives its value through the imposition of a &quot;profile&quot; upon a &quot;base&quot;. 5. Semantic structures incorporate conventional &quot;imagery&quot;, our ability to construe the same informational content in different ways.</Paragraph> <Paragraph position="2"> That meaning reduces to conceptualization (thesis 1), is characterized relative to cognitive domains (thesis 3), and incorporates conventional imagery (thesis 5) runs in stark contrast to the heavy emphasis placed on truth conditions and formalization by current computational linguistic approaches. We have attempted to tackle the informality of this ideational view of meaning, by addressing one particular basic cognitive domain, that of oriented three-dimensional space, and implement a restricted version of Langacker's process of conceptualization by means of conventional imagery. To verify the utility of the resulting conceptualization, we use the interpretations of spatial expressions so generated (the resulting images), to automatically construct a depictions of the scene. Theses 2, that prototypes should replace traditional objective categories, lies at the very heart of cognitive semantics \[Taylor1989\], and though it is widely accepted as true for semantic and most other linguistic categories, prototype theory is not conducive to rigorous formalization and has consequently been ignored by mainstream computational linguistics. Likewise our concern is with meaning variations that originate from different construals of the same information in the process of conventional imagery (thesis 5).</Paragraph> </Section> <Section position="10" start_page="304" end_page="304" type="metho"> <SectionTitle> IMAGERY AND ITS IMPLEMENTATION </SectionTitle> <Paragraph position="0"> This special technical use of imagery (not to be confused with the psychological term meaning the formation and manipulation mental images) refers to &quot;our amazing mental ability to &quot;structure&quot; or &quot;construe&quot;' a conceived situation in many alternate ways&quot; \[Langacker1988b\], as opposed to traditional semantic approaches whose concern is with informational content alone. Thus &quot;every conception reflects some particular construal of its content&quot;. Langacker identifies six important dimensions of imagery; in our semantic analysis of spatial expressions we are interested in just three of these: 1. level of specificity 2. scale and scope of predication 3. perspective The remainder of this section is a characterization of each of these dimensions and the consequences that their consideration has with respect to the design of a conceptual representation for spatial expressions.</Paragraph> </Section> <Section position="11" start_page="304" end_page="304" type="metho"> <SectionTitle> REPRESENTING 3-D SPACE </SectionTitle> <Paragraph position="0"> The basic cognitive domain relative to which the spatial meaning of projective prepositions is characterized, is structured three-dimensional space.</Paragraph> <Paragraph position="1"> In our system space is represented using an orthogonal axis system we refer to as the DCS (Domain Coordinate System). In the process of image construction conceptual objects will be constrained to locations described relative to the DCS. The DCS mirrors the speaker's perceptual assignment of axes to a scene, the x-axis extends from deictic left to deictic right, the y-axis from deictic front to deictic back, and the z-axis extends vertically.</Paragraph> </Section> <Section position="12" start_page="304" end_page="305" type="metho"> <SectionTitle> LEVEL OF SPECIFICITY </SectionTitle> <Paragraph position="0"> The level of specificity of conventional imagery addresses the issue of the degree of detail with which an entity is characterized. Specificity has already been mentioned in connection with the construction of the network of polysemous senses of a lexical item; on the other hand, concerning different lexical items, we can readily identify different spatial predications that are schematic with respect to each other. Consider the sentences below.</Paragraph> <Paragraph position="1"> (a) The chair is near the desk.</Paragraph> <Paragraph position="2"> (b) The chair is in front of the desk.</Paragraph> <Paragraph position="3"> (c) The chair is facing the desk.</Paragraph> <Paragraph position="4"> Sentence (a) simply predicates proximity; (b) predicates both proximity and a positioning of the LO relative to a particular side of the RO I ; lastly (c) predicates proximity and a relative positioning of the LO with respect to the RO, with the addi null tional anti-alignment of the fronl face normals of the two objects.</Paragraph> <Paragraph position="5"> Schematic contrast dictates the minimum degree of detail we must maintain in our computational representation of the conceptual reference and located objects. In sentences (a) the objects can be thought of as structureless points; in (b) the representation of the RO must incorporate the notion of sideness; and in (c) both the RO and LO are sided. We borrow Lang's conceptual representation of objects ZThe issue of which side of the reference object the located object is positioned with respect to is addressed as a consequence of the perspective dimension of conventional imagery termed object schemata \[Lang1993\], constructed within Bierwisch's and Lang's the two-level semantics \[Bierwisch and Lang1989\]. The object schema for a desk is: a max b vert c across</Paragraph> <Paragraph position="7"> In this first schema a, b and C/ label three orthogonal axes centered at the object, each of which can be instantiated by one or more dimensional assignment parameters (DAPs)2; al-a2, bl-b2 and c1-C/2 are corresponding half-axes. Each half axis is labelled either nil or with an intrinsic side (eg. i-fronl;). This representation is augmented with both a three-dimensional Cartesian coordinate which when assigned locates the conceptual schema relative to the DCS; and the values of the default extents for the object type along the axes a, b and C/.</Paragraph> <Paragraph position="8"> Imagery implies an imager, that is, the image exists in and with respect cognitive world of the speaker (by default) and this necessarily has important consequences. With respect to spatial language, issues pertaining to perspective, that is taking account of the imager, include the speaker's vantage point and orientation.</Paragraph> </Section> <Section position="13" start_page="305" end_page="305" type="metho"> <SectionTitle> ORIENTATION </SectionTitle> <Paragraph position="0"> The interpretation of some spatial expressions is dependent on assumptions as to the speaker's orientation with respect to the objects in the scene (eg. whether A is &quot;to the left of&quot; B in a scene, is dependent on the orientation of the speaker/viewer); other expressions are orientation independent such as &quot;above&quot; and &quot;below&quot; which implicitly refer to the downward pull of gravity (although in space verticality is speaker dependent).</Paragraph> <Paragraph position="1"> When an object schemata is characterized relative to the DCS it is both assigned a Cartesian position (as we show later), and its half-axes are assigned deictic sides according to their relative orientation with the observer. For example if a desk is positioned &quot;against the left wall&quot; as in figure 1 this would result an instantiated conceptual schema for the &quot;desk&quot; of: a max b vert c across</Paragraph> <Paragraph position="3"> are fundamental to the process of dimensional designation and and important where dimensional a~signment might result in a reorientation of the conceptual object (eg. &quot;the pole is high&quot;).</Paragraph> <Paragraph position="4"> Here al is the intrinsic left side but the deictic front of the desk.</Paragraph> </Section> <Section position="14" start_page="305" end_page="305" type="metho"> <SectionTitle> VANTAGE POINT </SectionTitle> <Paragraph position="0"> The speaker's vantage point is another factor that determines the interpretation of spatial expressions in a scene. The notions of deictic and intrinsic interpretations of projective prepositions can be accounted for purely by recognizing that in each the speaker adopts a different vantage point.</Paragraph> <Paragraph position="1"> For deictic interpretations the vantage point is the speaker's actual position. The vantage point for intrinsic interpretations is the functionally relevant position with respect to a reference object, for example, &quot;left of the desk&quot; under the intrinsic interpretation uses a vantage point that is directly in front of the desk (the typical configuration when a human uses a desk).</Paragraph> <Paragraph position="2"> The meaning of a projective preposition is conceptually represented as a spatial constraint on the conceptual schema of the located object which extends out from a particular side of a reference object, the precise nature of which we describe in the next subsection. In our system the lexicalized constraint is of the form of a two place predicate:</Paragraph> <Paragraph position="4"> cated object. The parameter side depends on the preposition. Thus the schematicity we observed in section is explicitly represented: (a) V is near X.</Paragraph> <Paragraph position="5"> < zonsprox X Y > Proximity constraint between X and Y. (b) Y is in front of X.</Paragraph> <Paragraph position="6"> < zoneprox X: front Y > Proximity and alignment of Y with front of X (c) Y is facing X.</Paragraph> <Paragraph position="7"> < zoneprox X:fron~ Y:back > Proximity, alignment and specific &quot;facing&quot; oriem</Paragraph> </Section> <Section position="15" start_page="305" end_page="306" type="metho"> <SectionTitle> SCOPE OF PREDICATION </SectionTitle> <Paragraph position="0"> Scope refers to exactly how much of a cognitive domain is included in the characterization. Minimally, the scope of an image for &quot;next to&quot; must encompass at least the reference and subject objects and some region of space separating them. We implement the spirit of this concept by realising the lexicalized constraint for a projective preposition as a potential field fixed at the reference object's position in the DCS 3, The proximity and directional nature of the constraint < zoneprox .. > is captured using a potential field P~, where:</Paragraph> <Paragraph position="2"/> <Paragraph position="4"> Here the x-axis points direction of the halfaxis of the particular side of the reference axis in the DCS; and in the case of &quot;in front of&quot; y is the perpendicular direction in the horizontal plane; (x0,y0) is the Cartesian coordinate of the reference object in the DCS, and lower the value of Pt~ for a location (x, y) for the located object the better the spatial constraint is satisfied. The minimum for the field can be quickly computed using gradual approximation \[3ramada et al.1988\]. The values of Kproz ~. Lproz ' ~r ' and Kdir,.~. are dependent on the located and reference objects and are set on the basis of scale considerations (see). Multiple spatial predications over an object is simply accommodated within the potential field model by linear addition of component fields.</Paragraph> </Section> <Section position="16" start_page="306" end_page="306" type="metho"> <SectionTitle> SCALE OF PREDICATION </SectionTitle> <Paragraph position="0"> The concept of the scale relates to the object dependency of the degree of proximity and directional constraint afforded by a preosition: where &quot;X is left of Y&quot;, and X and Y are houses, then the meaning of this predication would contrast with its meaning if X and Y were pieces of fruit. The concept of proximity and directional constraint predicated by &quot;left of&quot; is apparent in both cases, what differs is the scale relative to which it is characterized. null Scale effects are realised in the mechanism by which the constants of the potential field are set.</Paragraph> <Paragraph position="1"> For the potential field P~, the effect of the constants on the nature of the constraint are: :. K..o.,,~ Proportional to range of the possible separations of X and Y that would still satisfy the predication.</Paragraph> </Section> <Section position="17" start_page="306" end_page="306" type="metho"> <SectionTitle> 2. Lpro~,~ , </SectionTitle> <Paragraph position="0"> The default separation of X and Y.</Paragraph> <Paragraph position="1"> Proportional to the range of directions that would still satisfy the predication.</Paragraph> <Paragraph position="2"> Thus for a reference object that is a house Kp,.o~:,~, Lp,.o~,~, Kai,.~ r must all be considerably greater than for a piece of fruit. The precise values can only reasonably set as a result of some experimental investigation, currently Kp~o~, t~' and Lpro~ ,~ are linearly dependent on the sum of the extents of the reference and subject objects in the direction of spatial alignment; and Kdi~,~. on the perpendicular extent of the reference object in the plane of the constraint.</Paragraph> </Section> <Section position="18" start_page="306" end_page="306" type="metho"> <SectionTitle> GENERATING DEPICTIONS </SectionTitle> <Paragraph position="0"> After using gradual approximation to find the position of the minimum in the potential fields representing the spatial predications over a particular object, this point can be regarded as a probable interpretation. By tying each conceptual object to a graphical model, and interpreting the DCS as the viewer's perceptual axis system, conceptual interpretations can be rendered as scene depictions. Figure 2 illustrates one depiction of the cumulative interpretation of the following verbal description, in which all projective prepositions are viewed intrinsically 4.</Paragraph> <Paragraph position="1"> &quot;I am in a room. Against the left wall is a long desk. Against the back wall is a short desk.</Paragraph> <Paragraph position="2"> In front of the long desk is a chair. Another chair is to the left of the long desk. The chair in front of the desk is near the short desk.&quot;</Paragraph> </Section> <Section position="19" start_page="306" end_page="307" type="metho"> <SectionTitle> OTHER APPROACHES AND CLOSING REMARKS </SectionTitle> <Paragraph position="0"> Nearly all the work in recent years on computing the meanings of spatial prepositions stem from the prototype semantics of either Herskovits \[Herskovits1985\], \[Herskovits1986\] or Talmy \[Talmy1983\]. Schirra \[Schirra and Stopp1993\] adopts Herskovits' notion of a core meaning, and implements this as a typicality field. The ability to sum fields of different predications satisfies the compositionality requirement. Yet representational poverty exists with respect to the spatial and perceptual characteristics of the objects, as while directionality and proximity constraints are adequately captured for the intrinsic reference frame and set of objects, variation in the degree of constraint (for example, depending on the size of the reference object) and the potential for ambiguity arising from interpretations with respect to different reference frames are not accounted for.</Paragraph> <Section position="1" start_page="306" end_page="307" type="sub_section"> <SectionTitle> Underlying Kalita's </SectionTitle> <Paragraph position="0"> work \[Kalita and Badler1991\] is a conceptualization of the space around a reference object as six 4Natural language sentences are parsed to three branch quantifiers using a prolog DCG grammar, the logical predicates are the input to the cognitive semantic processor, the resulting conceptual representations are converted to depictions in by the depiction module . The cognitive semantic processor and the depiction module are implemented in Smalltalk/Objectworks bal description orthogonal rectangula~ projected regions (based upon an enclosing cuboid idealization of the object) due to Douglas \[Douglas and Novick1987\]. Using this model and following Talmy's work, the semantics of projective prepositions are lexicalized as geometric-relation schemas. Reference frame anabiguity is not addressed; directionality is too tightly restricted to one of the six rectangular regions, and proximity constraint is left to the &quot;underlying constraint satisfaction techniques and the use of a weight slot in the template for constraint representation&quot;.</Paragraph> <Paragraph position="1"> Within the framework of the LILOG project \[Maienborn1991\] Ewald Lang implemented the two-level approach to the semantics of dimensional adjectives in which the perceptual and dimensional properties of objects are conceptually represented as object schemata \[Bierwisch and Lang1989\]. Further developed for projective spatial predications, Lang's object schemata are capable of distinguishing deictic and intrinsic readings, though without explicit reference to a quantitative space (ie. actual scenes and observers) as in the case of Schirra and Kalita.</Paragraph> <Paragraph position="2"> Our system represents ~ first attempt, and very highly specialized implementation, of the conventional imagery process that is a component of the cognitive grammarian's view of linguistic semantics. Its performance, in terms of generating all possible interpretations, and the quality of the interpretations constitutes a significant advance on previous approaches.</Paragraph> </Section> </Section> class="xml-element"></Paper>