This is part 2 of a series of blog posts about modelling stems and principal part lists and covers the complexities in the notion of a lemma identifying lexical entries, specifically in the Pratt principal parts.

Before we get to the other principal parts beyond the first, there is a lot to be discussed just about the first part and its use as a lemma, identifying the lexical entry to which all the parts belong. In this post, we’ll start just looking at the presentation of lemmas in the Pratt list and in the next post move on to the other sources and the problems of merging multiple lists that may differ in choice of lemma for the same lexical entry.

The canonical lemma / first principal part is the present active (or middle) indicative first person singular of the verb but there are at least eight ways in which the first column in the Pratt principal parts table differs from this ideal.

1. Contract verbs

The present active indicative first person singular of a contract verbs like ἀγαπάω is, of course, not ἀγαπάω but ἀγαπῶ. The pre-contract version is often used (and is indeed used by Pratt) in lemmas and the first principal part so the stem vowel is explicit (as it’s necessary for generating other forms).

2. Base Verbs With a More Common Compound

Where a base verb gets its own entry but there is a more common compound, Pratt includes the latter in braces:

  • αἰνέω {ἐπαινέω}
  • ἀπατάω {ἐξαπατάω}
  • θνῄσκω {ἀποθνῄσκω}
  • ἵημι {ἀφιημι}
  • κτείνω {ἀποκτείνω}

Note that the other parts in this case are still given just for the base verb, even if that means they are not attested in Greek texts.

3. Compound Verbs

In some cases only one compound verb gets an entry, but the preverb is indicated in square brackets:

  • [ἀνα]λίσκω
  • [ἀν]οίγνυμι/[ἀν]οίγω
  • [ἀπ]αντάω
  • [ἀπο]κρίνομαι
  • [ἀπ]όλλυμι
  • [ἀπο]λογέομαι
  • [ἀφ]ικνέομαι
  • [δια]λέγομαι
  • [δια]νοέομαι
  • [δια]φθείρω
  • [δι]ηγέομαι
  • [ἐκ]πλήττω
  • [ἐπι]θυμέω
  • [ἐπι]μελ(έ)ομαι
  • [ἐπι]τηδεύω
  • [ἐπι]χειρέω
  • [καθ]εύδω
  • [καθ]ίζω
  • [κατα]δαρθάνω
  • [παρα]σκευάζω
  • [συλ]λέγω
  • [ὑπ]οπτεύω

It seems that compound verbs with a common base used for other compound verbs don’t get their own entries at all in Pratt and the base verb is to be referred to in that case. This is one example where bringing in metadata from Major’s list is potentially useful, in making sure common compound verbs can easily be looked up in their base verb form.

4. Multiple Present Stems Conjoined with Slashes

In these cases there are multiple alternative present (or more properly imperfective) stems conjoined with a slash.

  • [ἀν]οίγνυμι/[ἀν]οίγω
  • αὔξω/αὐξάνω
  • καίω/κάω
  • κλάω/κλαίω
  • μείγνυμι/μίγνυμι
  • οἴομαι/οἶμαι
  • σκέπτομαι/σκοπέω

While these could arguably be treated as separate lemmas (and hence lexical entries) there are two arguments against doing this: (1) the two forms given are really just alternative spellings; (2) the lexical entries converge in other parts.

5. Homographs That Differ In Other Parts

δέω has two senses that, while identical in form in the first part, differ in other parts.

6. Spelling Differences with Optional Letter in Parentheses

There are two cases where an optional epsilon is given in parentheses:

  • [ἐπι]μελ(έ)ομαι
  • οἰκτ(ε)ίρω

In some cases the spelling alternative continues into other parts.

7. Lexemes Where Other Lexemes are Merged In for Other Parts

These aren’t marked in the lemma itself but I’ve included them here as they represent a particular choice of lemma to group parts under. The actual parts from other lexemes are indicated by an asterisk in Pratt. Note that this is not the same as suppletion although arguably there is a fine line worth exploring in more detail at some point.

  • ἔρχομαι
  • ἐρωτάω
  • ἐσθίω
  • λέγω
  • πωλέω
  • ὠνέομαι

8. Lexemes Without An Imperfective Stem

Some words like οἶδα have a lemma which is from a part other than the first. While in some cases when this happens, the lexeme has been merged with another (see 7), this category covers the case where it hasn’t been.

Concluding Thoughts

We’ll see further issues when we look at the other lists and how to merge them but for now let’s discuss possible solutions to the issues seen already.

It is important to note that the information in the first column of the Pratt principle parts table (headed “present”) in the book is serving a number of distinct purposes:

  • providing an identifier for the entire row (what could properly be called the “lemma”)
  • providing the first principal part (and hence the present / imperfective stem)
  • providing additional information about the lexeme such as its preverb / base

By separating these out we have a much clearer way forward. The lemma proper can really be any unique identifier and it can be treated completely opaquely. The first principal part (or parts when there is more than one under a single lemma) can be a separate field. Finally, information such as preverb / base decomposition can be expressed in yet further separate fields. This keeps the first principal part free of extra characters and the lemma opaque.