Understanding the English preposition ‘ at ’ . Semantics and translation from a cognitive perspective

In this paper, which aims at bridging the gap between theoretical assumptions of semantics and practical problems of translation by means of cognitive linguistic tools, we will: a) ask whether prepositions are empty or ‘full’ lexical items, and, if full, see if it is possible to define their semantics; b) ask and try to understand what, exactly, ‘at’ (one of the most troublesome prepositions for foreign learners) means; c) try to address the frequently raised criticism that the representation of ‘at’ in dictionaries and other EFL materials is frequently unsystematic, vague, misleading, with mismatched and even inaccurate examples / translations, and make suggestions for the improvement of the representation / translation of the meaning of ‘at’. It should be noted that the suggestions presented in the conclusive part of this paper go beyond the concrete case of ‘at’, and are intended as being relevant for the representations of prepositional meaning in general. They are furthermore intended as a contribution towards a future detailed framework of cognitive linguistics seen as a field of scientific inquiry with a clearly defined subject, methodology and scope. Introduction: the task ‘At is perhaps the most troublesome preposition for foreign learners’ observes Lindstromberg (1997: 165) in his very detailed and insightful analysis of English prepositions. He explains the problems that foreign learners have with the English preposition ‘at’ by noting that ‘at is imprecise about the relation between Subject and Landmark ….[and that it] cannot be depicted’ (ibid., emphasis mine). While perfectly agreeing with Lindstromberg on the first point, i.e. about the semantics and usage of ‘at’ being extremely difficult to master for foreign learners, I strongly disagree with him on his second observation, relative to the semantics of the preposition ‘at’. In fact, in this paper, I would like to try to: a) show that the semantics of the English ‘at’ is quite precise, and b) suggest ways for improving the representation of ‘at’ in EFL materials. The paper is divided into three sections. In the first section, some general issues relative to prepositions are introduced. In the second section, we look at past attempts to define and formalise the meaning of the English preposition ‘at’, and then discuss them in the light of lexicographic and translation needs. Finally, in the third section, we try to combine theory and practice and propose a way for incorporating our findings into the context of EFL / ESL materials and, more generally, into the developing framework of cognitive semantics. Let us begin by introducing some notions that are important for our understanding of prepositional semantics, notions that we will later need for our attempt to accomplish the tasks set out above. 1. Preliminary notions 1.1. Prepositions or locatives The lexical items that fall within the category of prepositions are sometimes referred to as ‘locatives’. This is a reflection of the fact that prepositions can be studied from two different perspectives: a) the grammatical, and b) the ‘local’ (semantically driven). The main part of this paper focuses on the semantics of prepositions, but since there can be no exhaustive semantic account unless we also consider issues relative to the syntax-semantics interface, so let us start by observing the most important syntactic facts about prepositions. From the grammatical perspective, prepositions have, for a long time, been treated as merely “an annoying little surface peculiarity” (Jackendoff, 1973: 345). It is, then, encouraging from the perspective of the development of linguistic science, to note that the attitude towards this word class has changed considerably since the time when Jackendoff was able to challenge the linguistic community by saying that ‘people seem never to have taken prepositions seriously’ (ibid.). The main change has been given by the promotion of the lexical category of prepositions from the category of exocentric (i.e. non headed) constructions, into that of endocentric (i.e. headed) constructions. This means that from a syntactic point of view, prepositions are now treated on a par with verbs, nouns, and adjectives (they are now seen as projecting like other categories). Looking at things from a semantic perspective, it might be interesting to ask whether the recognition of a new syntactic role could be a symptom of the recognition of a semantic core of prepositions (since, when they were seen as simple ‘relators’, prepositions were almost completely devoid of meaning). Section 1.2. is aimed at answering this question. Talking about prepositions as a syntactic form class, we should spend a few words on the term itself. The name suggest that words belonging to this word class precede something (usually an NP) with which they form a phrase. However, in many languages (e.g. Japanese, Turkish, and Hindi) lexical items having the characteristics of prepositions appear regularly after the element they are related to (thus, postpositions). In English, too, they are found in positions other than relating two NPs (although we only study them here in this context). In order not to underestimate their syntactic (but perhaps also semantic) potential, we should mention cases in which they appear in adverbial position (for intransitive prepositions, as in ‘She ran across.’) or as subordinating conjunctions (for prepositions that take finite complements, as in ‘The day before she arrived’). For all the above reasons some people would rather have prepositions termed ‘adpositions’, but this name is still quite rare. Although, our analysis being semantic, it might seem more appropriate to use the term locative instead of preposition, given the familiarity of the latter term, we shall keep using it in the rest of this paper. We will conclude our remarks about syntax by noting that in both languages under consideration for the purposes of this paper (i.e. English and Croatian), prepositions are, in traditional grammar books, listed under the heading of the morphologically invariable parts of speech, forming a closed class. Having reviewed some general issues relative to prepositions within the syntactic framework, let us now turn to the core problem of this paper: the semantics of prepositions as a word class and then, more concretely, the semantics of the English ‘at’. 1.2. The meaning of the word class As already noted, linguistics has, for some time now, been familiar with the idea that syntactic categories express certain semantic traits which are common for all members of a given syntactic category (e.g. Talmy, 1983, 2000; Slobin, 1985, Levin and Pinker, 1991). Is it then possible to establish a ‘general meaning’ for prepositions, i.e. the word-class as such? If we take a look at one of the traditional grammar reference books, we read: “a preposition expresses a relation between two entities, one being that represented by the prepositional complement.” (Quirk, 1985: 673). If prepositions are, by definition, relational words, then in order to understand the nature of their meaning, i.e. of the type of relation they can establish, we need to stop for a moment and think about the sort of things they put into relation. Herskovits (1986: 7) notes that the simplest type of prepositional spatial expression is composed of three constituents, i.e. the preposition and two noun phrases (NP), as in ‘the spider (is) on the wall’ The two NP-s are referred to in the literature by various names (‘theme’, ‘located entity’, ‘located object’, ‘spatial entity’ … for the first NP, and ‘reference object’, ‘reference entity’, ‘localiser’, ‘landmark’ … for the second NP). The terminology adopted in this paper is: Figure (abbreviated as ‘F’) for the first NP, i.e. the object being located, and Ground (abbreviated as ‘G’) for the second NP, i.e. the object in reference to which F is being located. The notions of Figure and Ground were originally described in Gestalt psychology, but their application in linguistics stems from Talmy (1983), who characterised them as follows: “The Figure is a moving or conceptually movable object whose site, path, or orientation is conceived as a variable the particular value of which is the salient issue. The Ground is a reference object (itself having a stationary setting within a reference frame) with respect to which the Figure’s site, path, or orientation receives characterisation” (Talmy, 1983: 232). Given that preposition seems to relate F’s location with respect to G (F’s location being static in the case of locational prepositions and dynamic in the case of motional ones), we might easily be led to conclude that the relation established by a preposition (as word class) has a locational or topological nature. But let us contrast: ‘F (is) above G.’ vs. ‘G (is) above F.’ While it would be possible to view the two sentences as differing only in expressing a different (reversed) location of F and G (thus the preposition having a pure topological character) I would like to hypothesise that the difference between F and G in the two sentences is much deeper, with the semantic contrast residing in a differently profiled conceptual content (cf. Langacker, 1987), with the profiling being functionally motivated. Here is what I mean.