Getting the Right Handles on the Greek Perfect

At SBL 2013 there was a session devoted to a description of the Greek perfect. Basically, there were three positions and they had been published previously. Stanley Porter argued that the perfect was a third aspect category in Greek with a meaning of “stative”. Buist Fanning argued that the perfect entailed a complete action with existing results/relevance. Constatine Campbell argued that the perfect was an imperfective aspect.

In one sense, Fanning provided a succinct warning or potential falsification of Campbell’s view by mentioning that imperfective was an insufficient parameter for dealing with the Greek perfect. An ‘imperfective only’ view would better fit the semantics of a verb like ἀποθνῄσκω ‘he is dying’ since an imperfective by definition does not include the end-point of the event and a person cannot say of a corpse ‘he is dying’. Campbell’s position would not semantically differentiate ‘he is dying’ from ‘he is dead’ or ‘he has died.’ (Technically, not discussed at the session, the converse of Fanning’s example has a potential occurrence. One may say of someone dying but still alive, ‘he is dead,’ meaning that death is a foregone conclusion and inevitable. But one may not say of a corpse, ‘he is dying.’) Fanning’s point was the only really testable and falsifiable point in the discussions, most of the discussion revolved around integration with published literature and nomenclature. Porter did not want to have a label ‘perfect,’ perhaps because that would entail some kind of anterior event within the grammatical category. Fanning, of course, spent time showing that such a logical entailment is necessary, especially in some of the active perfects.

One thread in the discussions raised the matching of the meaning of the perfect with its morphology. Campbell pointed out that Greek has four morphological stem categories perfective (aorist), imperfective, perfect, and future. He defended his position of having only two aspects (perfective and imperfective) by claiming that a three aspect system (Fanning: perfective, imperfective, perfect; Porter: perfective, imperfective, stative) does not match Greek morphological categories on a one-to-one correspondence, so there is no theoretical reason that would prevent a two aspect system. I would like to to show how the Greek perfect morphology applies to the Greek perfect. It works out quite elegantly. The Greek perfect is an aspectual category and its morphology reinforces the viewpoint that it is not only an ‘imperfective’ like Campbell, nor is it simply a third aspect category like Porter’s stative, but that it is an aspect that has a perfective element inside it, like Fanning. (I will not deal with the future. Although it is a fourth morphological verbal stem category, it is not an aspect and needs separate treatment.)

The perfect is a complex, fused category that includes two ordered parameters that may be labelled {+perfective, +imperfective}. The perfective parameter explains why the perfect includes a complete action. Think of a sentence like ἐμήμεκα “I have vomited.” There is a complete event, the vomiting. The person was not about to vomit or in the process but is viewed as having vomited. In addition, the imperfective parameter in the perfect is a continuing viewpoint that gives the existing relevance. On the other hand, an aorist ἥμεσεν ‘he vomited’ does not make any present claim or implication, but ἐμήμεκεν ‘he has vomited’ implies that someone is now sick, or is now free from something foul in their stomach, or that something needs to be cleaned up. There is an ongoing relevance. Since the existing relevance does not change, it is amenable to a term ‘stative’, but the perfect active includes the perfective event its viewpoint. One could use the label ‘stative’ to describe the ongoing relevance, as long as the ‘completeness’ parameter was also included.

If we look at morphology we find a very impressive correlation between form and meaning. The perfect active forms have a reduplication and a “κ” (kappa) in the perfect stem. Thus, there are two morphological markers in the perfect active, that iconically match the two parameters that Greek scholars have always seen in the category. (‘Iconically’ means that the form mimics the meaning in some way.)

A kappa is a marker that is recognized from such core verbs as ἔδωκα “I gave” and ἔθηκα “I placed.” As very common Greek verbs, these ‘kappa-aorists’ are preserving an archaic feature of Greek morphology that ceased to be used in classical and Koine Greek. The kappa may be said to be a fossil or remnant of a perfective marker. This kappa is close to the core stem and sits approximately where a ‘sigma‘ would sit in stem formation for aorists. Such a ‘core’ stem marking is appropriate for an ordered parameter in the perfect that is semantically prior to any imperfectivity or stativeness.

The reduplication marked in the perfect may be considered an iconic marking of extension and imperfectivity. (Imperfectivity is a continuative viewpoint that excludes the endpoint of the action and an open, unchanging situation may be considered a subset of imperfectivity). Many imperfective stems add pieces for the imperfective stem (i.e. ‘continuative’ stem), for example λαμβάνειν adds a ‘ν’ and –θνῃσκειν adds an ‘σκ’. More importantly for this discussion, διδόναι and τιθέναι specifically use reduplication for marking the stem that is {+imperfective (stative)}.

So the perfect active has two morphological markers, ‘κ‘ and reduplication, and these can be directly linked to two parameters ‘perfectivity’ and ‘imperfectivity’.

The perfect middle, on the other hand, does not have a ‘κ᾽. That might seem like an anomaly, but on reflection it, too, fits better than might have been expected. In fact, this is quite an unexpected, delightful correspondence between form and meaning. The middle and passives are more prototypically pictured as a result, a state, without picturing the event that produced the result. Consider a schema like a governor who signs a bill. When we say ‘the governor has signed the bill’ we include the governor’s signing within our mental representation. This is similar to ‘the boy has vomited’- we picture the action of the vomiting in our mental representation even while we consider the implications that the perfect tense implies. However, with a statement like ‘the bill has been signed’ we do not include a picture of the signing ceremony in our mental representation but we focus on the result, the state. A potential image of ‘the bill has been signed’ might include a picture of the signature on a document, but not the action of the governor. What I am saying here is that an active perfect of an ‘action verb’ forces us to keep the action itself current in our mental representation while the middle-passive allows one to focus on the resulting state. Somewhat surprisingly the Greek morphology of the perfect appears to iconically reflect this difference of focus. We may provide a little bit of this Greek distinction in English because we could be said to have two present perfects, depending on the helping verb, ‘he has died’ and ‘he is dead’, and ‘it has been signed’ versus ‘it is signed.’ Greek is not far from this because some of the potential Greek morphology of the perfect has been carried by a periphrastic structure where some of the perfect subjunctives, optatives and imperatives are filled in by a form of the verb ‘be’ and the perfect participle.

I am not saying that the categories of the active perfect and middle perfect have different semantic aspects, I treat them as the same category, but they do have potentially different orientations as a result of interacting with ‘voice.’ It is important to keep both the +perfective feature and +imperfective feature in the perfect because we need to keep ‘he is dead’ (perfective) distinct from ‘he is dying’ (imperfective). Both the statements ‘he has died’ and ‘he is dead’ have a complete starting point along with its present implications, although the second form ‘he is dead’ has a focus on the state in English. So both perfect active and perfect middle would have the parameters of ‘perfectivity’ and ‘imperfectivity’ included with a complex, fused category in Greek. However, the reduced focus on the causative action in the perfect middle has apparently led to a category marking in Greek that “dropped”(or better: did not add) the ‘κ‘ marker.

Technically, one could choose to name this complex category with a term like ‘perfect’ because of the focus with the active, or a term like ‘stative,’ because of the focus with the middle. The Greeks called the category παρακείμενος ‘lying there.’ Both terms ‘perfect’ and ‘stative’ have problems with extra baggage in English. The perfect in English is a tense, while the Greek perfect infinitive, subjunctive, and participle are describing an aspect. However, the Greek perfect indicative is also a tense, like English, and includes a location in the present, a present perfect, and a location in the past, a past perfect (pluperfect, plusquamperfectum). The term ‘stative’ raises some problems because the term is commonly used for describing a kind of verbal action and is used for describing a class of lexical meaning rather than a grammatical category. The simple solution in English is to keep the term ‘perfect’ so that a connection with Greek-English reference tools is maintained. Using the term ‘perfect’ would assume that in the indicative a perfect was a short way of saying present-perfect. The simple solution in Greek is to use the terminology that they have used for 2000 years. The perfect indicative is παρακείμενος and the pluperfect is ὑπερσυντέλικος.

In any case, regardless of the chosen metalanguage, the dual complexity of the Greek perfect is mirrored in its morphology in a remarkably consistent way by including two distinct marking systems. I would suspect that someone discussing the development of Greek morphology out of Proto-Indoeuropean may have already stated this 100 years ago. That will have to wait for a month until I am near a research library. Now I need to board a flight.