Variation within a single class: achievements
The traditional account of Aktionsart classes assumes that the four classes are homogeneous cross-linguistically. On this account, verbs which have similar meanings should have similar aspectual properties from one language to another, since they will belong to the same Aktionsart class. In this section we will see that the picture is not as simple as this: cross-linguistic investigation reveals some fairly significant differences in the temporal properties associated with verbs of the same Aktionsart class in different languages.
In order to show this, we’ll explore a classic achievement verb, die, dis cussed by Botne (2003). Recall that achievements are instantaneous or minimally lasting occurrences, in which the actor passes from one state to another: realizing the truth, losing/finding something, noticing something, and so on. Dying is just about as instantaneous an event as they come. One moment we’re alive, the next moment we’re dead: blink and you might miss it. Dying is thus something that happens in an absolute minimum of time. Of course, the lead-up to the moment of death might be protracted, but someone who’s near to death is still, quite clearly, alive. The instantaneous nature of die in English is reflected by its incompatibility with in the midst of: one cannot, in English, be in the midst of dying. Other classic achievement verbs show the same restriction: *I’m in the midst of realizing the truth/losing my wallet/noticing a problem.
Botne proposes that achievement verbs can, in fact, potentially encode durative temporal phases. This doesn’t mean that their traditional classification as punctual is wrong; just that we need to adopt a more detailed view of their meaning. According to Botne, achievement verbs aren’t limited to simply expressing a single punctual culmination point. Instead, some achievement verbs may have a more complex temporal structure, in which the central point is surrounded by one or two subsidiary phases, a preliminary (‘onset’) phase and a ‘coda’ phase . . . It is this tripartite structure which constitutes ‘the underlying fabric of the event’ (Botne 2003: 236).
Botne claims that all languages have a verb which expresses the instantaneous transition from life to death. Where languages differ is in the temporal phases surrounding this central nucleus. Let’s begin our cross linguistic survey by looking at English die. Botne claims that die in English consists of two phases: the punctual nucleus, and a durative onset phase.
His main evidence for this claim is that, unlike some other achievement verbs, die can appear in the progressive, as in (57):
Given that achievements are described as punctual or instantaneous, this is a somewhat unexpected situation. Some other achievement verbs, like buy, can appear in the progressive. Many, however, resist progressive contexts:
Consider the force of the progressive be dying. This refers not to the transition moment between life and death itself, but to the lead-up to this moment; to be dying typically means to be in the last stages of life. Someone who is dying is still alive. (This explains the occurrence of die with the present continuous that we noted in (50b) above. Win, as exemplified in (50a), has exactly the same explanation.)
Unlike achievements, activities and accomplishments are durative and, as a result, freely appear in the progressive. But there is an important difference in the effect of the progressive on achievements and activities/ accomplishments. Consider the entailment structure of the activity weep and the accomplishment recover (as in recover from an illness):
QUESTION Compare the entailments of is buying the paper, is opening the door and is realizing the truth. Do they pattern like die?
With die, the progressive does not entail that any amount of dying has already occurred – the opposite of the case with activities and accomplishments. Botne concludes from this that our conceptualization of dying consists of two phases, a durative onset phase, which is what the progressive targets, and a punctual nucleus.
Botne (2003: 240) states that Arabic, Hausa (Afro-Asiatic; Nigeria) and French are like English in that their die verbs consist of a durative onset phase preceding the nucleus. He calls this inceptive type coding. In Egyptian Arabic, for example, the root mwt ‘die’ takes the progressive pre fix bi-, as do non-achievement verbs:
The possibility of the progressive here is explained by the hypothesis that mwt is not simply an instantaneous state-transition event; the verb’s temporal structure also includes an anterior phase which can have durative structure.
Another type of temporal structure found in die verbs is what Botne calls resultative-type encoding. We will illustrate this from Japanese. The Japanese counterpart to the English progressive is the so-called -te iru construction, illustrated in (63):
This construction consists of a gerundive form of the verb (glossed GER) and the auxiliary iru, an imperfective form of ‘be’. When applied to sinu, ‘die’, the -te iru construction does not produce a ‘progressive’ meaning. Instead, it refers to the state of death that results from the instantaneous transitional moment:
Botne claims that Japanese has no way to refer to the onset phase of dying. Rather, all one can say is that someone is ‘about to die’, or ‘appears about to die’, as in (65).
All this says is that the moment of death is imminent – not that it is already ‘in progress’.
Lastly, we will discuss the purest type of temporal structure for die, in which the verb only expresses the instantaneous nucleus. An example of this structure is Assiniboine (Siouan; Dakota, USA). Assiniboine t’a can only have a punctual reading, referring to a non-extended point of time (66a). This contrasts with other verbs, such as activity verbs, which also have a progressive reading without any modification to the root (66b).
However, t’a is compatible with neither of these suffixes.
How, then, does one express the inceptive meaning encoded in the English progressive? To convey this meaning, Assiniboine uses the auxiliary verb aya ‘become’, as in (68):
Botne concludes from these facts that t’a is best thought of as only encoding the nucleus transition.
QUESTION Can you think of any other ways of analysing the facts?
What is the upshot of this survey? Botne’s discussion shows that the verb meanings typically classed as achievements have a more complex temporality than was originally assumed. We cannot simply describe achievements as instantaneous transition points and leave it at that. What is punctual about achievement verbs like find, die, notice, or recognize is their nucleus; in addition to this nucleus, they may well express a durative onset or coda.
