7

If I have a verb in ancient greek, how can I find its root?

For instance, if I have

λείπω
νέω
ἔμαθον
μανθάνω

how can I do to know that, respectively, these verbs have

 λιπ-/λειπ-/λοιπ-
 νευ- (<*νεϜ-)
 μαθ-

as roots? Do have I study by heart all verbs with their roots?

My grammar explains to me how to form the present theme, but not how can I have roots. For example, it explains to me that βλέπω has its verbal root equal to its present theme, but, I suppose, if I find a verb I do not know, how do I know that its verbal theme is equal to its verbal root?

For polythematic verbs, like λέγω (= I say) and ὁράω, I know it is necessary to study their paradigms.

Of course, I can look for all my verbs in a good dictionary, but I would like to understand if it is possible to go back to the first person (-ω or -μι) with a logical reasoning and not going by trial and error.

  • 1
    Is "type it into the Perseus morphology tool" a valid answer? I've never found a better one. – Draconis Aug 24 at 17:11
  • @Draconis Very interesting, thank you, but it does not answer to my question, sorry. – Onner Irotsab Aug 24 at 20:25
  • To clarify: are you looking for patterns how to derive the aorist root from the present tense (for example)? Greek has six principal parts, and they aren't always derivable from each other, but there are patterns that are helpful in memorization – b a Aug 25 at 20:31
  • @ba Grammarians know that λείπω has λιπ-/λειπ-/λοιπ- as roots, respectively, and I wonder how they know it, that's to say, how they find this information. I am trying to investigate inner mechanisms of greek. Do you mean paradimg as "pattern", i.e. βλώσκω, μολοῦμαι, ἔμολον, μέμβλωκα or λύω, λύσω, ἔλυσα, λέλυκα, λέλυμαι, ἐλύθην, λυθήσομαι are patterns? – Onner Irotsab Aug 26 at 9:48
  • 1
    @ba I am interesting in: from λείπω how to deduce ἔλιπον (and not, following regular formation for weak aorist, ἔλειψα (ε-λειπ-σα > ε-λειπ-σα > ε-λειψ-α) that derives from λείβω) and vice-versa. I would like to understand if there is a logic/scientific way to do this and not (only) a mnemonic way. I have spoken of verbal roots because, since the endings are always the same (main and hystorical endings), the whole conjugation depends on the verbal roots and the usual phonetic phenomena between vowels, diphthongs and consonants. – Onner Irotsab Aug 26 at 12:10
6

Welcome to the site! I'm afraid there isn't exactly one way to rule them all. But there are various phonological rules by which you can guess the roots of a significant number of verbs.

  • For example, -(i)sk, -nu, and -an are common present suffixes, so cut them off if you want to find the root. The -an- suffix is in the present manthanô (root math-); -nu is in deiknumi (root deik- or dek- or dok-); -sk- is in heuriskô (root heur-). There are various other suffix that aren't part of the root, like the common -eu- and other vowel clusters in pisteuô, dêloô.

  • Gemination of the final consonant is often due to a present suffix (invisible -y), too. Cf. ballô (< bal-y-ô), root bal-.

  • The infix -n- is a common in present stems as well. It can be rendered as mu (perhaps also as gamma) due to assimilation with the consonant after it: it is a mu before labials, gamma before velars, I believe. Examples are manthanô, root math-, and lambanô, root lab-.

  • Reduplication (semi-repetition of the first syllable) can be a present praefix, and is often a perfect suffix. E.g. present didômi, root do-, perfect leluka, root lu-.

Etc. So removing all affixes is often a good thing to try in case you want to go from present stem to root/aorist. Some affixes are used in all classical forms of the verb but not in nouns or other words from the same root, so you might only see the true root outside verbs. Of course the converse is also possible (nominal affixes not present in verbs).

Then there is the phaenomenon of Ablaut: in Proto-Indo-European, which is also Greek's ancestor, many roots had three potential manifestations, 1.) a "zero grade" where nothing is added to the root, e.g. lip, 2.) an "e grade", e.g. leip, and 3.) an o grade, e.g. loip. So you can try removing an e or o when present, or adding one when not present in case you're going from root to present stem. Other examples are peithô, root pith-, and pet-o-mai, root pt- (aorist e-pt-o-mên, perfect pe-pot-ê-mai).

A nice example is ker-an-nu-mi "knead" (as in English ceramic). It has -nu, and it has -an-—or is that really only -n-, since many other forms of the verb have the a without the n (aorist e-kera-sa)? The verb seems to show Ablaut, too, as we have perfect ke-kra-mai. Liquids (m, l, n, r) surrounded by too many consonants are often vocalised (turn into or acquire a vowel sound) in Greek; in Attic-Ionic dialects, this is usually a, so that might be where the a came from in the perfect, which was then perhaps generalised and also used in places where it wasn't 'necessary', as it often happens in language.

I'm sure there are other rules of thumb that apply to some other verbs. But there will be many where you probably can't find a rule that would apply, alas. So you will need to learn all verbs by heart, in theory; but you can save yourself a lot of work by trying these rules of thumb; they are never guaranteed, but I believe they will provide you with a high chance of success for quite few verbs.

5

Greek verbs have six principal parts, meaning that to be able to conjugate a verb in all of its tenses, you need to know all six different roots with their conjugations. Sometimes the roots used in each principal part are identical (e.g. with λύω), sometimes similar (e.g. λείπω), and sometimes (as you already acknowledge) entirely different (e.g. λέγω). Here are some rules that can help with the middle category of similar, but modified, roots, to determine which parts of the present root of the verb are absent in the aorist root (this is essentially an expansion of Cerberus' point about removing affixes).

Certain sounds at the end (i.e. before the first person suffix -ω) or middle of the present tense are deleted in the aorist. I take these from Wikipedia's article on Ancient Greek present progressive markers, which goes into more detail.

  1. Roots ending with the various reflexes of consonant + *y: -ζ- (derived from a root of τ,δ,θ + y), -σσ- (from κ,γ,χ + y; Attic reflex is -ττ-), -πτ- (from β,π,φ,ψ + y), -ιν- (from ν + y), -ιρ- (from ρ + y)

    • Examples: θαυμάζω (aorist ἐθαύμασα), πράσσω (aor. ἔπραξα = ἔ-πραγ-σα), χαίρω (aor. ἐχάρην)
  2. Ending with -σκ-

    • Examples: εὑρίσκω (aor. ηὗρον = ἔ-ευρ-ον), πάσχω (> *παθσκω, aor. ἔπαθον)
  3. Ending with -ν-, -νέ-; or of the pattern with two nasals -NCάν-, where C is a consonant and N the corresponding nasal (μ, ν, γ)

    • Examples: δάκνω (aor. ἔδακον), λαμβάνω (aor. ἔλαβον)
  4. Ending with -έ-, or containing -ε- before a vowel earlier in the root

    • Examples: λείπω (aor. ἔλιπον), δοκέω (aor. ἔδοξα = ἔ-δοκ-σα)
  5. Starting with a reduplication of the first consonant

    • Examples: μιμνήσκω (aor. ἔμνησα), βιβρώσκω (aor. ἔβρων = ἔ-βρω-ον) (both also ending in -σκ-)

Outside of the domain of present progressive markers are irregularities due to other sound changes. Your example of νέω, for instance, is irregular due to a regular sound rule (namely, deletion of ϝ). ἔχω has an irregular aorist (ἔσχον) for more specific historical reasons. Neither are predictable from the root itself.

Finally, the rules also have exceptions. For example: διδάσκω is reduplicated, but doesn't lose the reduplication in the aorist (ἐδίδαξα) as most reduplicated verbs do. There is no algorithm that guarantees a way to correctly guess the aorist from the present root (hence why they are different principal parts), but being able to recognize the patterns can help you memorize when irregular conjugations usually occur.

  • 1
    To the first category can be added -λλ- from -λy-, e.g. βάλλω, ἔβαλον. – TKR Aug 27 at 21:11
  • This is a bit more systematic than my answer. Out of curiosity, are you suggesting the -e- in dokeô is from Ablaut (like the internal -e- in peithô)? I don't remember exactly, but I always thought it was a different -e-? – Cerberus Aug 27 at 21:16
  • @Cerberus I don't know. The Wikipedia article I quoted them from groups them both together, so I just followed that – b a Aug 27 at 21:54
  • @Cerberus The two -e-s have nothing in common; the first is ablaut, the second is a suffix (-ey-). Wiki is misleading to group them together. – TKR Aug 27 at 23:47
  • @TKR: OK I suspected as much. I don't think I've ever seen Ablaut manifested after the coda. – Cerberus Aug 28 at 1:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.