4.1 Figure and ground, trajector and landmark: early research into prepositions
When we look at an object in our environment, we single it out as a perceptually prominent figure standing out from the ground. The same principle of prominence is valid in the structure of language. For example, in locative relations like in The book is on the table the book is conceptualized as the figure.
In Chapter 1, section 1.2, we mentioned the notion of gestalt as a basis for the categorization of objects. Besides such perceptual principles as the laws of proximity, closure and continuation, the gestalt psychologists were very interested in how our visual and auditory input is organized in terms of the prominence of the different parts. To take an example from the auditory domain, when we listen to a piano concert we can easily make out the part played by the piano as being more prominent than the accompaniment of the orchestra. Similarly in the visual mode of perception, when we watch someone doing a high jump our eyes will follow the movements of the high-jumper rather than rest on the crossbar or the ground. In this chapter we will have a closer look at how the findings of the gestalt psychologists affect the study of language.
Figure/ground segregation and locative relations
To start with, consider the picture of the well-known face/vase illusion shown in Figure 4.1. You will notice that of the two possibilities of perceiving the picture (as two faces or as a vase) you can only see one at a time. Still, you can easily switch between the two ways of looking at the picture, especially after longer inspection. What lies behind our inability to see both the vase and the faces at the same time is a phenomenon called figure/ground segregation. This notion was first introduced into psychology by the Danish psychologist Rubin almost a century ago and later integrated into the more comprehensive framework of perceptual organization by the gestalt psychologists (see also section 1.2 of Chapter 1).
Examining our visual experience when looking at Figure 4.1 more carefully, we notice that what we single out as figure seems to have special properties. The figure has form or shape whereas the ground is formless and the shared contour seems to belong to the figure. Besides shape and contour the figure seems to have other thing-like qualities such as structure and coherence, whereas the ground is unstructured, shapeless and uniform. The figure appears to lie in front of the ground which extends more or less continuously behind it. All in all, the figure is perceived as being more prominent than the ground, and psychological research has shown that it is more likely to be identified and remembered, and to be associated with meaning, feeling and aesthetic values.1
While a prolonged inspection of Figure 4.1 clearly confirms these aspects of the perceptual prominence of the picture, it does not really explain why at one time the vase and at another the two faces assume a figure-like quality. In other words, it would be interesting to know what factors govern the choice of the figure. Since the objective input to our visual system does not change, this choice remains entirely up to the observer; yet it seems not very plausible that it is just a matter of personal taste or whim.
To answer this question one must realize that the face/vase illusion is of course an example of a very special nature, because it allows for what is called a ‘figure/ground reversal’. Most visual scenes which we encounter in our everyday lives are of a different kind in that they suggest a particular figure/ground segregation. Thus in the scene depicted in Figure 4.2 the obvious entity to be chosen as figure is the book, while the table will be given ground status, unless a special and quite unnatural effort is made to see the table as figure.
According to the gestalt psychologists the principle of Pragnanz, which we have already introduced in Chapter 1, section 1.2, plays a major part in assigning the status of figure to certain parts of a visual scene. Unlike the table, the book in Figure 4.2 readily fulfils such gestalt principles as the principle of closure (its contours are closed) and of continuation (it is an uninterrupted whole). In addition, the relatively small area of dark colour and its balanced proportions attract our attention and make the book a more likely candidate for the figure than the table. Finally, the book seems more likely to be moved around than the table (for example to be picked up), and this may also contribute to our natural choice of figure (Talmy 1978).
This last aspect is even more important when we turn to examples where an object is depicted in motion as the balloon is in Figure 4.3. Looking at the picture we will not hesitate to regard the balloon as figure and the house as ground because, being conceived as a moving object, the balloon seems to be much more prominent than the house. From a linguistic point of view it is interesting to consider h o w the selection of figure and ground and the relation the two elements have to each other in our two Examples is rendered in words. Describing the situation depicted in Figure 4.2 we would probably say that the book is on the table, thus claiming a specific locative relationship for figure and ground. Similarly we would assume that the balloon in Figure 4.3 is above the house, or when we think that the balloon must in fact be moving, we might prefer the description The balloon is flying over the house. Such a relationship between balloon and house would take into account that the position of the balloon changes in time as suggested by Figure 4.4.
What these two examples show is that the relationship between figure and ground can be seen in terms of locative relations, which are usually rendered by prepositions; or to put it the other way round, the meanings’ of locative prepositions can be understood as a figure/ground relationship. It is this second view, the prospect of being able to explain linguistic expressions such as prepositions in terms of figure and ground, that has caught the attention of cognitive linguists.
Image schema, trajector, landmark and path: some fundamental notions of prepositional analysis
Before exploring the link between prepositions and the figure/ground contrast more closely we should perhaps remind ourselves of how prepositional meanings are understood in cognitive linguistics. As already explained in Chapter 2, section 2.4, locative relations like —OVER— and —UNDER — , —UP— and — DOWN — , —IN— and —OUT— are regarded as image schemas, i.e. simple and basic cognitive structures which are derived from our everyday interaction with the world. The idea is that by experiencing for example many instances of things-over-things we have acquired some sort of cognitive pattern or schema of the — OVER— relationship which we can apply to other instances of this locative relation. The schema which has thus developed is obviously less concrete than the rich prototype categories of objects and organisms discussed in earlier chapters. However, as the term suggests, an image schema is not just an abstract semantic principle, but should be understood as a mental picture which is more elementary than both concrete categories and abstract principles. This is something to keep in mind when we now look at pictorial representations used for locative image schemas in cognitive linguistics.
Let us start with a graphic representation of the image schema of —OVER— which underlies our example The balloon is flying over the house (Figure 4.5).
When we compare Figure 4.5 with Figure 4.4, it is obvious that the representation in Figure 4.5 is less concrete. The balloon (which we identified as the figure) and the house (the ground) are now represented by symbols. The reason is of course that the image schema —OVER— also applies to other objects apart from balloons and houses (e.g. The bird is flying over the tree or The kite is flying over the hill). The symbol for the figure (the circle) appears several times to indicate that what is represented is a process, something that has a dynamic quality. Each circle stands for a different temporal stage of this process. Moving from stage to stage the figure follows a path. Since the path of a bullet or missile can be understood as its trajectory, the figure is called ‘trajector’. The ground functions as a reference point for orientation and is therefore called ‘landmark’. This means that the notions ‘trajector’ and ‘landmark’ are specific manifestations of the more widely applicable notions of ‘figure’ and ‘ground’. This use of trajector and landmark has been generalized in cognitive linguistics, so trajector stands for the figure or most prominent element in any relational structure (and is therefore indicated by very bold lines), whereas landmark refers to the other entity in a relation.
Surveying the diagram and the explanations, one might get the impression that this representation is nothing more than a pictorial illustration of our model sentence or a set of closely related sentences. Such a view does not take into account that all the elements involved can be modified in various ways. Trajector and landmark may vary in size and shape, the trajector can be in contact with the landmark or it can even be part of the landmark, as will be illustrated in the following sections. The important thing is that all these variations can be derived f r om one and the same relationship between trajector and landmark, in our case the -OVER— relationship. In other words, the trajector/landmark approach promises to provide a description of the meaning of a preposition); which reveals the relation between its various senses. Such a unified cognitive description of the various prepositional meanings is in strong contrast with earlier views which regarded them as an array of unconnected senses.
Unified descriptions need a starting point, a core, a schema which can be regarded as ‘central’ (we try to avoid the term ‘prototypical’ because it suggests a rich category structure which is not characteristic of image schemas). Here we follow Brugman’s (1981) analysis of —OVER— as summarized by Lakoff (1987)3 and Lindner’s (1982) study of the meaning of—OUT— and —UP— in verb-particle constructions.
Over, out and up: the central schemas and some elaborations
In accordance with Brugman and Lakoff we accept Figure 4.5 as the central schema for —OVER—. It consists of a trajector moving along a path that is above the landmark and goes from one end of the landmark to the other and beyond. As we have seen, the trajector may stand for a balloon, a plane, a bird, a kite, while the landmark may represent a house, a wall, a tree, a hill, etc. In addition, there are cases where the landmark is unspecified as in the sentence The plane flew over. Here it is the position of the speaker or conceptualizer that functions as a landmark, so the example The plane flew over is similar in meaning to The plane flew over me/us.
Let us next have a look at the central schemas of the prepositions out and up. Figure 4.6 provides the graphic representation of this central schema as used in the sentence She went out (of the room).
As the diagram shows, in the first stage the trajector is included in the space occupied by the landmark, which can represent an object like a room or which can be unspecified, implying ‘She went out from where we are’. Viewed in isolation, this initial stage roughly corresponds to what is denoted by the preposition in and it can therefore be regarded as the image schema of — IN —. The specific aspect in the schema —OUT— is that the trajectory moves from being included within the boundaries of the landmark to a location where the two are completely detached from each other. This is indicated in the diagram by the three positions of the trajector.
Compared with —OVER— and —OUT — , the image schema for —UP — , which is illustrated in Figure 4.7, is a more difficult case. Here the path of the trajector has a vertical direction, the landmark is only relevant as far as its vertical extension is concerned. This is why the landmark is only represented by a vertical line in the diagram even where it is specified as in The boy climbed up the wall. (After all, when you climb up a wall you do not think of its thickness, but you are only interested in its vertical dimension, in the vertical distance that is involved.) This rather abstract notion of landmark may be more difficult to grasp and to accept than other types of landmark, but it makes it easier to imagine that the landmark is unspecified, as in sentences like The rocket went up.
Let us pause at this point and review the characteristics that are shared by schemas —OVER—, —OUT— and —UP—. From the evidence presented so far we can conclude that all three central schemas represent cognitive configurations consisting of three elements and their interrelations, namely
(a) a trajector, which moves along
(b) a path, and is seen as being related to
(c) a landmark.
As we have seen, the relationship of trajector, path and landmark is sufficient for a rough distinction of the meanings of the three prepositions. To account for more specific meanings or uses we have to consider the variations which trajector, path and landmark can undergo. Such variants which only (specify ‘certain components of a schema, but do riot diverge from its general configuration, are called elaborations.
Figure 4.8 illustrates some of the elaborations observed for the image schema —OVER — . Variant (a) depicts a very common case, which applies for many verbs of motion, namely contact between trajector and landmark. Variant (b) shows that the trajector can assume a size and shape which is similar to that of the landmark and be located quite close to it so that the trajector actually covers the landmark. Finally, variant (c) illustrates what should probably be regarded as a marginal case, the fusion of trajector and landmark. The wall that collapses (or at least its upper part) is the trajectory that is moved along a path and it is at the same time the landmark which acts as reference point for the motion.
Figure 4.9 contains another set of elaborations, this time related to the central schema for —OUT —. What most of them have in common is that the variation concerns the initial stage, the way in which the trajector is (or is not) integrated in the landmarks This initial stage is contrasted with the final stage because it is this contrast between the initial and the final stages of the process in which the meaning of —OUT— manifests itself; providing the middle stage would not add any important information.
By directing the attention to selected stages of the image schema, the analysis of —OUT— has pointed the way to how we might approach another puzzling facet of prepositional use: the fact that essentially dynamic prepositions like over, out and up are used to render meanings that might at least superficially be regarded as ‘static’. Compare the following examples:
(1) Hang the painting over the chimney.
(2) She stays out.
(3) She lived three floors up.
To start with the first example, the position of the painting as described in the sentence can be easily understood as the final stage of a process of moving the picture up into its position. This is represented in Figure 4.10.
Similarly, sentences (2) and (3) can be explained as the final stages of the path of the —OUT— and —UP— schemas. Compare Figure 4.11.
Metaphorical extensions
So far we have seen how a fairly general schema of a locative relation can be employed to explain a considerable variety of uses of prepositions in terms of elaborations. It has also become clear that the links between the various elaborations are based on a similarity relation to the central schema which they specify in a number of respects. To some extent this relationship may be regarded as being analogous to the similarity between the members of a cognitive category of objects or organisms, a view which has been advocated by some cognitive linguists.
Although the notion of schema has considerable explanatory power, it cannot account for all uses of the three prepositions. Where this is the case, that is, wherever the meaning of a preposition differs from the central schema in such a way that it cannot be explained by reference to the schema alone, we speak of an extension of the schema. Since an immense number of extensions exist for all prepositions, we cannot do justice to all of them. We will therefore single out a few instances of the preposition over which illustrate the kind of processes involved in extensions. The examples are taken from Lakoff (1987; examples 3 to 5 have been slightly adapted):
(1) She has a strange power over me.
(2) Sam was passed over for promotion.
(3) Harry still hasn’t got over his divorce.
(4) The government was overthrown.
(5) The match is over.
How can these uses be explained? In Chapter 3 we pointed out that next to basic level categories, it is image schemas that provide metaphors with their source models. These image schemas which, as we have seen in this chapter, mainly represent locative relations assist us in conceptualizing abstract categories and cognitive models. A by-now-familiar example is the +CONTAINER+ metaphor based on the —BE IN— relation, which is involved .in our understanding of ideas and communication in general, and of course in the conceptualization of many emotions, most notably anger. Similar kinds of metaphorical mappings are at work in the case of over.
The conceptual metaphor which explains the use of over in (1) above is derived from the basic locative experience —UP/DOWN—: + CONTROL IS UP/LACK OF CONTROL IS DOWN+. According to Lakoff and Johnson (1980: 15) this is only one example of a whole host of so-called ‘orientational metaphors’ which map basic experiences from the spatial domain onto abstract cognitive models. (Other examples are +HAPPY IS UP/SAD IS DOWN +, +CONSCIOUS IS UP/UNCONSCIOUS IS DOWN+ and +HEALTH AND LIFE ARE UP/SICKNESS AND DEATH ARE DOWN+.) Example (1) is explained very easily if we interpret it as a cognitive extension of the structure of the basic image schema in Figure 4.5 from the cognitive model of PHYSICAL SPACE to the model of POWER RELATIONS. Metaphorically speaking, a person who has the power to exert control over another person is the trajector which is in a higher position than the landmark. This metaphor is so familiar to us that more often than not we are not even aware of the fact that we use locative expressions when we talk of power relations (e.g. have control over someone, be under someone’s control, be on top of the situation, high society, the lower classes).
In example (2) the same metaphor +CONTROL IS UP/LACK OF CONTROL IS DOWN+ is supplemented by a second metaphor, namely +CHOOSING IS TOUCHING+, which is related to the contact elaboration of the schema (Figure 4.8(a)). The image conjured up by the expression being passed over for promotion is that of a person who is in control of other persons’ status. From this superior vantage point the person picks out particular people from the many who are below him or her. Since the —OVER— schema normally indicates that there is no contact between trajector and landmark, it can be concluded that in the example sentence, Sam was not among those chosen.
In example (3), Harry still hasn’t got over his divorce; again two metaphors are at work. The more general one is the metaphor +LIFE IS A JOURNEY+, which, as we have seen in section 3.1, maps the structure of the event category JOURNEY onto the abstract category LIFE. It lies in the nature of journeys that the traveller’s progress is sometimes impeded by obstacles such as rivers, gorges and of course hills and mountains, and this is where the second, more specific, metaphor +PROBLEMS IN LIFE ARE OBSTACLES IN A JOURNEY + comes into play. If bad times, problems or setbacks in one’s life are seen as vertical obstacles like mountains in one’s life journey, one has to get over them in order to move on in life (cf. also He is over the hill). Viewed from the complementary perspective of the schema of—OVER— (cf. Figure 4.5) this means that Harry is the trajector, who, in order to pursue his path (= a journey = his life) must get over the landmark (= an obstacle = the divorce).
The sentence The government was overthrown in (4) is related to the earlier example The fence fell over mentioned and explained in Figure 4.8(c). Lakoff argues that what lies behind (4) is again the metaphor +CONTROL IS UP+. ‘Before the event takes place, the government is in control (metaphorically upright), and afterwards it is not in control (metaphorically, it has fallen over)’ (Lakoff 1987:439). One could probably add that another metaphor which may roughly be characterized as + PROPER FUNCTIONING IS UP+ is also involved. From our everyday experience we know that plants, animals, persons and indeed objects like bookshelves, houses or spires can only fulfil their respective functions in a satisfactory way if they keep their upright position. Lying on the ground, or still worse being scattered around, is a clear sign for an organism or object which is either currently not active or severely damaged and this may contribute to the rich image evoked by the expression overthrow the government. A closely related metaphor that was mentioned above is +HEALTH AND LIFE ARE UP/SICKNESS AND DEATH ARE DOWN+.
Finally, example (5) The match is over is based on a structural mapping from the cognitive model SPACE onto the model TIME. The event category MATCH is understood as an extended locative landmark. The more specific underlying metaphor therefore is +AN EVENT is AN AREA+ and the general picture is that we travel along a path through a game just as we may travel over a field or a bridge. In terms of the schema of —OVER—, the persons travelling are the trajector (which is not expressed on the linguistic surface of this expression) and the game functions as a metaphorical landmark. When the travellers reach the boundaries of the landmark, the field and the event are over.
In conclusion, the discussion in this chapter has given a first glimpse of how the principle of prominence, which underlies the figure/ground distinction, is at work in the structure of language. So far we have confined ourselves to a few rather simple examples from the area of prepositional relations and their non-locative extensions, but more about the principle of prominence will be said in the next two sections. The crucial notions that have proved valuable for the description of locative relations are ‘trajector’ and ‘landmark’, ‘path’, ‘schema’, ‘central schema’, ‘elaboration’ and ‘extension’.
As far as the relation to previous chapters is concerned this section has taken up some issues that have already been addressed before, for example locative relations from Chapter 2, section 2.4 and conceptual metaphors from Chapters 3, sections 3.1 and 3.2. While these issues were still very much related to the lexical aspects of language, we will now probe a little deeper into the role of prominence in grammatical structures.