Lexicalization patterns in motion verbs
Come and go differ in the type of path they express: in come, the path is oriented towards the deictic centre, whereas in go it need not be. Path, the route traversed by the object in motion, is one of the major elements of a motion situation. In English, motion verbs often do not express path in the verb root, as come and go do. Mostly, the path element is expressed in a preposition or ‘particle’, like in, out, away, along, down, through etc., with the verb root expressing the manner of the motion. Thus, English has a wide range of verb roots indicating different manners of motion: crawl, run, roll, walk, skip, fly, float, stroll, tumble and so on. In themselves, these verbs do not convey anything about the path along which the motion takes place: to express this, it is necessary to specify the path using a directional expression:
In an influential study, Talmy (1985) compared how different languages lexicalize the four elements of motion, path, manner and figure (the moving object), which he took to be the essential components described by motion verbs. He was particularly concerned to see how the manner and path components are shared between the verb root itself and elements like prepositions and particles. Talmy coined the term satellite to refer to this latter type of structure. Satellites are ‘certain immediate constituents of a verb root other than inflections, auxiliaries or nominal arguments’ (1985: 102). Off, out of the room and over the Alps in (17) are all satellites.
Talmy hypothesized the existence of three basic combinations of the possible components of the motion event, depending on whether the basic motion element is paired with manner, path or figure. Languages differ in which of these three options predominates. As we will see, English is biased towards the first, but also illustrates the second:
The third lexicalization pattern, motion + figure, is the least represented in English. A language that exemplifies it generously is Atsugewi (Hokan; northern California):
QUESTION Talmy proposes that rain and snow exemplify this third pattern in English. Can you think of any other examples?
Talmy proposed a major typological division between what he called verb framed and satellite-framed languages. This division concerns whether the path component is lexicalized in the verb root itself or in a satellite element.
Germanic languages like English and German are principally satellite framed: most verbs of motion are not like enter or exit (both of them loan words in English). Instead, most Germanic motion verbs express the manner in which the motion occurred, and any specification about the path has to be introduced in a separate locative expression. Consider, for example, the German sentences (from Brecht 1967: 81) and their English translations in (19)–(20):
English walk denotes a particular manner of motion and does not say anything about the path the motion took: this is conveyed in the satellite prepositional phrases in an exclusive street (19b) and beside the shop windows (20b). The German original has exactly the same structure. The verbs promenieren and schreiten (past tense schritt) both mean ‘walk’, expressing both the fact and manner of motion in a single form. Like English, German encodes the path in a satellite, consisting of a prepositional phrase (in einer vornehmen Straße/ an den Auslagefenstern). Slavic, Celtic and Finno-Ugric languages are also satellite-framed.
Romance languages, however, are characteristically verb-framed: the path is specified in the verb root itself. Here is an example from Spanish:
As the literal translation ‘exit’ makes clear, the verb salen inherently expresses the path element ‘out of’. Greek, Semitic, Turkic, Basque, Korean and Japanese are all verb-framed languages like Romance. Here is the same sentence as (21) in Basque (isolate; Spain and south-west France):
Notice how the most natural English translation – ‘all the bees fly out of the hive’ – doesn’t reflect the literal structure of the original. It is worth emphasizing that statements about such and such a language being verb or satellite framed do not mean that every motion verb in the language is of the appropriate type; it is a question of which type is most charac teristic of the motion expressions in the language. Talmy defines ‘characteris tic’ as meaning (i) that the verb-type is the one found in colloquial, not literary, language; (ii) that it occurs frequently, and (iii) that it is pervasive, meaning that a wide range of different types of motion are expressed by it.
QUESTION Assemble as long a list as possible of English motion verbs, and note whether they include a Path component. Are there any where it is hard to decide? Is Talmy’s classification of English as a satellite framed language justified?
Talmy illustrated this with the following selection of Spanish motion expressions, all of which show the verb-framing characteristic of the language. Comparison with the English translations shows how systematically the two languages diverge: English always expresses manner in the verb, and path in a satellite, while Spanish expresses path in the verb, and manner in a satellite (flotando).
Note that the indication of the path isn’t limited to the verb in Spanish: all the sentences contain satellites which convey additional path-related information. So in (23a) the verb entró supplies the information that the path is an inwards one, whereas the satellite a la cueva tells us that it had the cave as its goal. In the right context it would also be possible to say La botella entró desde la cueva (fl otando) ‘the bottle floated in from the cave’. This shows that the verb itself expresses a different path element from the one mentioned in the prepositional phrase. This is also clear from (23e) and (23f), where the difference between the verbs corresponds to a difference in path, in spite of the identical prepositional phrase.
As well as verb- and satellite-framed languages, some linguists claim that there is a third type, equipollent languages. This is a type in which both path and manner are treated in the same way by the language’s morpho syntax (see Slobin 2004, 2006). Most equipollent languages are ones with serial verbs, i.e. verb complexes consisting of several independent verbs, each making a separate semantic contribution, as in the following sentence from Papiamentu (Afro-Iberian creole; Netherlands Antilles):
The verb phrase contains a sequence of equally central and morphosyntactically equivalent verb morphemes ‘fly, go’, which does not admit any sort of verb–satellite distinction on Talmy’s criteria.
Talmy’s typology has been widely discussed, and the distinction between verb- and satellite-framing is often invoked as a way of characterizing how different languages distribute motion information in the clause. It has not gone unchallenged, however. We will consider two types of criticism.
The first typically hinges on Talmy’s notion of characteristic motion expression, mentioned above. As we have seen, languages usually contain other types of lexicalization pattern than the one reflected by their classification as verb- or satellite-framed. These categories are idealizations which capture what Talmy takes as the predominant, most basic type of lexicalization pattern in the language. This leaves it open to other scholars to probe whether, and how far, the idealization is justified. Kopecka (2006: 97), for example, claims that French ‘does not correspond to a consistent type within Talmy’s typology and furthermore exhibits a greater variety of lexicalization patterns than had previously been recognized’. This is because there is a large number of basic motion expressions in which the path is expressed by a prefix, a satellite element. This conflicts with the status of French as a verb-framed language in Talmy’s scheme. Some of the many possible examples are accourir ‘run to’ and atterir ‘land, touch down’, formed with the prefix a(d), s’envoler ‘fly away’ and s’enfuir ‘run away’, formed with the prefix en- and parcourir ‘run all over’, formed with the prefix par (see Kopecka 2006: 86 for more examples). A similar criticism is made for Spanish by Cuartero Otal (2006). This type of criticism does not undo the distinction between verb- and satellite-framing, but simply questions its status as a language-wide phenomenon. If many languages initially taken as exemplars of one type prove to be mixed, Talmy’s principal typological conclusion – that languages typically display a single lexicalization pattern – will be disproven.
The second criticism questions the legitimacy of the very category ‘motion verb’. Talmy-style analyses take this as a basic semantic class and as the site of the major typological distinction between verb- and satellite- framing languages. Concentrating on French, Cadiot et al. (2006) argue that it is a mistake to see ‘displacement’ – physical motion in space – as the basic component of the meaning of many of the verbs relevant to Talmy’s analysis. This suggests that Talmy’s typology overemphasizes a single aspect of what is actually an intricate and multifaceted array of meanings. Cadiot et al. claim that the traditional way in which we describe the meaning of French motion verbs is basically flawed. It is a mistake, they suggest, to see displacement as the most central aspect of the sense of motion verbs, even if they are obviously often used to refer to motion events. They claim that the assumption that motion in space is semantically basic is untrue to the experiential grounding of language. Talmy’s distinction of path, manner and figure as fundamental components of the motion scenario ignores the fact that human beings do not experience motion in abstracted, ‘geometrized’ terms; as they put it, the abstract framework of space in the Talmy tradition ‘is neutral with respect to any practical engagement’ (2006: 187). We do not simply move from point A to B in a particular manner, but do so with aims and intentions, in a way that involves many types of subjective, perceptual and qualitative factors which are ignored by Talmy’s analysis. Cadiot et al. argue that these additional factors reveal themselves in the numerous non-spatial, non physical uses of motion verbs, which have to be taken as metaphorical or otherwise non-literal in Talmy-style approaches. For Cadiot et al., a unified analysis of these various uses is possible which does not privilege displacement as the key notion. This unified analysis avoids postulating a literal, basic motion use and a set of non-literal semantic extensions from it. Instead, they claim it is possible to discern aspects of meaning common to the so-called ‘literal’ and ‘metaphorical’ uses alike which reflect the distinctive subjective character of the experience of motion expressed by the verbs.
The French verb tomber ‘fall’ is an example. For Talmy, tomber would count fundamentally as a verb of manner of motion. But Cadiot et al. propose that other aspects are equally important, specifically the aspects of verticality, suddenness, non-control and surprise. These elements are all features of the human experience of things that fall. For Cadiot et al., it is illegitimate to treat these as secondary. Indeed, their presence is revealed in uses of the verb usually considered as metaphorical, such as the following (among others):
Uses like these do not involve displacement. Instead, they foreground aspects of the meaning of tomber which Cadiot et al. claim are always present even in ‘literal’ uses referring to motion. Ignoring them in a way that treats motion as the principal ‘core’ of the verb’s meaning is, they suggest, an important distortion of the semantics of tomber. One can, indeed, use tomber in a way exactly parallel to the English ‘s/he fell down/ over’, and this is a use which need not encode any motion in the sense of displacement in space: one can fall without changing one’s spatial position in any significant sense. Instead, what has changed when a subject falls over is the degree of their control over their own body. In this instance, motion of the type privileged by Talmy, displacement on a path between distinct points, is not an important aspect of the verb’s meaning. To make displacement central to the semantic analysis of tomber is to ‘ignore dimensions bound to the subject, which are not necessarily associated with an actual displacement, but rather with a change that is perceived from inside and outside the subject, and that cannot be reduced to an external trajectory in a topological space’ (Cadiot et al. 2006: 182). Again, these ‘subject-bound’ dimensions of meaning are revealed in a use of the verb usually treated as metaphorical or extended, and thus as derived from some more basic meaning. The expression tu tombes bien, literally ‘you fall well’, expresses the idea ‘you’re just in time’: a meaning in which abstract motion is significantly less important than the more subjective dimensions of suddenness or surprise.
QUESTION Consider English fall, assembling as many representative examples as possible. How far is abstract, Talmy-style motion a central part of its meaning? Is an analysis in Cadiot’s terms attractive?
The verb monter, glossed as ‘go up’, is exactly analogous. Here too it would be a mistake to treat motion as the central aspect of the verb’s meaning. Indeed, the gloss ‘go up’ obscures the fact that monter can have transitive uses which have little to do with spatial verticality, as Cadiot et al. explain (2006: 194):
It is essential to notice the dimension of intentional programming or the anticipation of a terminal point, which is more readable in the ‘assembly/put together’ uses (monter un kit ‘to assemble/put together a kit’, or even monter une maison ‘to build a house’, where the construction process is considered to be inherently programmed) or in the constitution uses (monter un projet ‘to set up a project’). We therefore see an inherent telicity or programmed aiming at the center of the meaning of monter. . .
Verbs like tomber and monter, then, are semantically more complex than their simple treatment as motion verbs implies. Cadiot et al. criticize investigators in the wake of Talmy for their privileging of motion, which leads them to artificially introduce it as a component of the meaning of these verbs in many cases where it is not in fact relevant. A particular case is the metaphorical expression la route monte, an exact French equivalent of the English ‘the road goes up’. This is usually explained as involving metaphorical motion based on the personification of the road, or as representing the mobile point of view of a subject following the road uphill. Yet this use is better understood, Cadiot et al. claim, as instantiating the semantic feature of ‘anticipation of a terminal point’ referred to in the passage just quoted. The road does not in any sense move: indeed, precisely the point of a road as opposed, say, to an escalator, is that it is not itself in motion. To introduce motion into the semantic analysis of la route monte is therefore unreasonable. The use of monter is explained by what Cadiot et al. see as a permanent feature of its semantics, the notion of ‘anticipation of a terminal point’. To say that the road ‘goes up’ is to register the difference in verticality between its initial and terminal points, not to attribute motion to it in any way. A similar case would be the verb sortir ‘come out’, which, in French just as in English, applies to many cases where there’s no actual physical motion, like la photo est bien sortie ‘the photo came out well’.
Cadiot et al. do not deny that real, physical motion between spatial points is often a part of the meanings of verbs like monter, tomber and sortir. But they do not believe that it should be privileged as the unique or determinative aspect of their semantics. In their opinion, analysis in terms of the motion of a figure on a path is insufficiently focused on the embodied, subjective qualities of our experience of these actions, and reflects an overly abstract, conceptual approach to meaning. This analysis springs from a very different understanding of meaning from Talmy’s. Talmy’s approach involves abstracting from the multiplicity of uses of motion verbs and concentrating on just one aspect of their meaning. By contrast, Cadiot et al. resist the instinct to abstract, believing that there is a basic mistake involved in taking displacement as the central aspect of the semantics of verbs like monter, tomber and sortir. Instead, they emphasize how the meanings of these verbs reflect subjective, qualitative dimensions of experience which are not easily reduced to configurations of paths and figures. In doing so, they offer a more holistic, but considerably more complicated analysis.
The two styles of analysis can coexist. Talmy can always claim that his analysis does not have to be taken as the end of the story about motion verbs’ meaning. The verb/satellite distinction only targets those aspects of the verbs’ meaning which are relevant to displacement between points, and nothing in it precludes the more subjective, qualitative approach advocated by Cadiot et al. If displacement is not an important part of the meaning of many motion verbs, this in itself does not challenge the typo logical division between verb- and satellite-framing; it simply deepens our appreciation of the semantic complexity of the verbs in question.
