The following is a Q/A interaction about some of the implications stemming from the application of an assymetrical view of markedness to the Greek verbal system. I have only taken a small piece so far, but some excellent questions have arisen regarding the next steps in the application. The exchange is repeated below to help illustrate moving beyond the broad application to a finer, more nuanced consideration of certain areas. My responses are indicated in blue. The general point is that one is not finished with an explanation when the aorist is claimed as default. This is a 50,000 foot claim that needs to be carried forward based on more specific factors. One of those is voice, another is distribution. What if a form does not occur in all of the tenses? What about the idiosyncrasies of certain lexemes? Markedness is an organizational framework. It should serve the analyst, not vice versa.
I’m thinking that “markedness” is indeed a very useful analytic notion, and that your “asymmetrical” is the right sort. I do have problems in some instances with the apparent “either/or” implication of the binary opposition notion. My problems come to the surface with your reference to a common notion that the aorist and perfective aspect should be viewed as the “default” or “unmarked” while the imperfective aspect and present & imperfect tenses should be viewed as the “marked” form — marked for imperfectivity. I wonder whether this really doesn’t depend upon the Aktionsart of the particular verb; clearly there are verbs like εἶναι that have no aorist, but I think there are also verbs like τύπτειν which never displays an aorist in the GNT or LXX, although the aorist is certainly present in older Greek. Might there not be verbs for which the imperfective is the “default” form?
SER: This is where “choice implies meaning” comes in. Markedness is just an organizational framework. As the members of the set change, one may have to adjust the organizational framework. If “aorist” is not a choice, then there is nothing to be learned from calling it a default. One would need to look at the usage and distribution of TUPTEIN, and create a new framework suited to the actual data. My post is assuming a full set, and is painting in very broad strokes, too broad in the case of TUPTEIN. I am trying to provide a general framework. The more closely one is working with a specific set, say verbs of motion, one may need to adapt based on the new data.
I’m even more concerned with this problem of “markedness” with regard to active and middle-passive voice morphology. Clearly the “active” forms are default, unmarked — I would say — for subject-affectedness, while the middle-passive forms are clearly marked for subject-affectedness. BUT: there are several categories of verbs for which it appears that the middle-passive form is the default; for intransitive verbs of process I think we’d have to say that these verbs are perhaps “unmarked for causality” while the active morphology for such verbs is “marked for causality.” Does that make sense? Or is it a reasonable way to use the term “markedness”? A verb of interest here is ἵσταμαι and its aorist ἔστην: that intransitive form is, I think, the “default” form, while ἵστημι and its aorist ἔστησα are marked for causality. There’s a problem of how we understand the second-aorist forms, especially the -HN/HS/H forms, which seem generally to be intransitive, although they take “active” endings.
SER: One of the things that I find most useful about markedness is the ability to nest one set within another. For instance, you mention the sets of active, middle and passive. If one takes the active as the most basic, the others are explained by how they differ. One could also take the middle as the default, and say that the active marks causality, perhaps. Whichever route you take, you would then be able to look at the middles as a subset, and begin to group them in meaningful ways, based on Aktionsart, semantic domain, by whatever prove to be meaningful factors. I will be doing this with the present forms themselves, once I finally get there. We will subdivide the presents into the default–reference to current time or “- remoteness” for the aspect only crowd. Decker’s work in Mark shows that this accounts for 34% of the forms, or 54% if you add what he considers to be temporally unrestricted. I will be looking at the remainder of the data, breaking them into groups based on what they mark as present. Nesting like this has tremendous explanatory power, but is often overlooked or ignored.
BUT: there are several categories of verbs for which it appears that the middle-passive form is the default
This is partially true, that is, some verbs do not have a active form, but that does not mean that the active still isn’t default within the Markedness system. The middle/passive form is default only in the sense that its normal, but not in the sense that its unmarked.
Though I know Carl wouldn’t use this terminology, his discussions of voice on his homepage suggest quite clearly that voice, linguistically speaking, is semantic role agreement.
If that’s the case, then the middle/passive is still marked since those verbs which take it normally are semantically distinct from those that do not. What we need is a broad study that examines these verbs more closely for making generalizations about the semantics of these verbs.
It would be interesting to compare the situations in Greek and in Russian, as their verb systems are remarkably similar.
In Russian there are many verbs in which the imperfective is morphologically basic and the perfective is derived from it by prefixation, and there are also many verbs in which the perfective is basic and the imperfective is derived by suffixation. (There are also secondary imperfectives, formed from a basic imperfective with a prefix to make it perfective and a suffix to turn it back into imperfective. And there are secondary perfectives formed from a perfective with a suffix to make it imperfective and a prefix to make it perfective again – and possibly even some with a second suffix to turn it back yet again into an imperfective.) I see traces of the same system in Greek, e.g. the -an in verbs like lambano looks very like one of the Russian imperfectivising suffixes and suggests that in verbs like this the aorist stem lab- etc. is the more basic form.
I’m not sure quite how these morphological changes relate to markedness, but the morphological base forms were presumably the original defaults.
Peter, Thanks for the comment. I recall Mike Aubrey making similar comments about Russian, but my skills are too rusty to add anything. Generally speaking (at least according to Foley and Van Valin), the perfective aspect is typically the least marked, and thus the most frequently occurring. Levinsohn provides a tidy presentation of their claims in “Discourse Features”, section 10.2 on “natural prominence.” Having said this, one might end up reorganizing the system somewhat based on the constraints of Russian. Remember, it is an organizational system that organizes the set, but there ought to be typological principles that are followed. This is the thrust of Foley and Van Valin’s discussion.