I just finished adding the principal parts to the DCC ancient Greek core vocabulary list, something I meant to do last summer, but which got lost in the shuffle. So that’s done, and up. Phew. Anybody who has tried to learn ancient Greek knows what a big hurdle the principal parts are: absolutely essential, but a beastly task of brute memorization. I am here to say that, as one who focuses more on Latin than on Greek, I have to re-learn some of them on a regular basis if I want to read (or teach) Greek well. This is not the fun, life-affirming, profound, aesthetically enriching part of Greek. This is the boot camp, the weight-lifting one must do to get there.
The idea behind principal parts is to put in your hands, and hopefully in your brain, all the different stems of a verb, so that (theoretically) any declined form can be derived from, or traced back to, one of them. But of course it’s not quite that simple.
On the one hand, some verb forms and related things are extremely common, but not really directly derivable from the principal parts as they are traditionally presented. εἰκός, for example, is a very common participial form meaning “likely, plausible” that is not immediately apparent from the principal parts of ἔοικα. It’s in the dictionary, of course, but somewhat buried in the entry on ἔοικα.
On the other hand, many Greek verbs have principal parts whose stems are only very rarely employed. πέφασμαι, for example, is a perfect tense principal part of a very common verb, φαίνω. But forms derived from it are rare. πέφαγκα, another perfect form listed by Smyth among the “principal” parts is very rare indeed, with only seven attestations in the TLG, almost all of those from late antique grammarians and lexica. I guarantee you will never encounter it outside a grammar book.
Part of the problem here is that our apparatus for learning ancient Greek is largely derived from big, comprehensive, scientific grammars of the 19th century, and thus have a tendency to completism, rather than the conveying of what is most essential. This is a general problem that does not only affect the issue of principal parts.
Enter into this picture the database, specifically the TLG and its lemmatizer tool. This is the tool that attempts to determine from what dictionary head word (or lemma), a given form derives. I have complained elsewhere about the impotence of existing lemmatizers when it comes to determining the meaning of homographs–forms that are spelled the same but derive from different lemmas, or forms derived from a single lemma, but which could have more than one grammatical function. This is a serious and as yet unsolved problem when it comes to asking a computer to analyze a given chunk of Greek or Latin. And the homograph problem also substantially compromises frequency data based on machine-analyzed large corpora of Greek and Latin.
But one thing at which the lemmatizers are extraordinarily good–theoretically flawless– is telling how many occurrences of a certain word form there are in a given corpus. And by examining that data you can get in most cases a very accurate picture of how common are the forms derived from a particular stem or principle part in a Greek verb. In other words, the TLG Lemma Search (which is what I have been working with in making the principal parts lists for our site), helps us see more clearly than has ever been possible which principal parts of each verb are the most important, and which very common forms lie slightly outside the traditional lists of principal parts. It has the potential to make principal parts lists far more informative and helpful to the language learner even than the information found in Smyth, LSJ, or any of the current textbooks.
I can think of a couple ways in which TLG lemmatizer data could be used to enhance the presentation of Greek principal parts. One could, for example, have a second list of, say, the five most statistically common forms of a given verb. In the case of πάρειμι, for example, that would be the following (with the total raw occurrences in TLG as of today):
παρόντος (8587), παρόν (5406), παρόντα (4920), παρόντων (4442), παρόντι (3451)
In fact the top 10 or so are all participial. παρών παροῦσα παρόν: that’s what I call a principal part!
Another way to do it would be to print in bold the principal part from which the most forms derive, or even use a couple different font sizes to reflect how commonly used each principal part is. For σῴζω, save, the figures are (roughly) as follows σῴζω (8600) σώσω (1300), ἔσωσα (5500), σέσωκα (400), σέσωσμαι (700), ἐσώθην (8800). Interesting to see the aorist passive stem beat out the present stem. The top vote-getters in terms of forms are σωθῆναι, ἔσωθεν, σώζεται/σῴζεται, σῶσαι, and σῶσον.
People who are better at Greek and spend more time with large corpora and their analysis than I do have probably thought of all this long ago, and there may be some principal parts lists that incorporate some of this data. If so, I would love to hear about it.
Before closing I should give a huge thank you to Prof. Stephen Nimis from Miami University of Ohio and his collaborator Evan Hayes, whose principal parts list in their edition of Lucian’s A True Story (soon to be re-published on our site with extra features) was of great assistance as I was making our list. And I should mention here also the crucial help I have had all along with our Greek list from the great Wilfred Major, of Louisiana State University.