Recently, I saw the following tweet:

The noun ἅδρυνσις ‘[the] coming to maturity’ is a derivation of the verb ἁδρύνω ‘to ripen, mature’, which is in turn a derivation of the adjective ἁδρός ‘thick, stout, bulky’. The adjective can refer to both people, animals, and fruit or crops.

You can display the derivation as follows:

Adjective ἁδρ   -ός
Verb ἁδρ -ύν
Noun ἅδρ -υν -σις

Although it is not unheard of that I get excited by Greek vocabulary, this particular word, ἅδρυνσις, caught my attention. Why is that? Normally, it is the unexpected realization of a vocalic nasal or laryngeal that excites me, but this time it was the cluster -νσ-. As far as I was aware, it should not be there: either the nu (ν) or sigma (σ) normally dissolves in such clusters between vowels (with or without lengthening of the preceding vowel).

Skimming through Buck (1933; 150-152), you learn that there are two results of this cluster between vowels (you can consult Buck’s Comparative Grammar for yourself):1

  1. If the cluster is original, that is from Indo-European, Proto-Greek or a loan, the sigma disappears with compensatory lengthening (the vowel becomes long) of the preceding vowel:
    • The sigmatic aorist (1st singular) loses its sigma (and then becomes pseudo-sigmatic, obviously): ἔ-κρῐν-σα > ἔ-κρῑν-α ‘I decided’; or ἔ-νεμ-σα > ἔ-νειμ-α ‘I stayed’.
  2. If the sigma in the cluster is not original, because historically it used to be a *t, the nu disappears with compensatory lengthening of the preceding vowel:
    • *-ti̯- (the iota is a consonant before a vowel: yod) becomes -σ-: παντ-ι̯α > πανσ-α > πᾶσα ‘all, everything’ (nominative singular feminine).2
    • *-ti- (the iota is a vowel) becomes -σι-: *φερ-οντι > *φερ-ονσι > φέρ-ουσι ‘they carry, bear’ (Attic present indicative 3rd plural).3
    • Or *dental stop (τ, δ or θ) + σ, where the sigma assibilates the dental stop: *φεροντ-σι > *φερον-σι > φερου-σι ‘to the bearing (men)’ (Attic dative plural masculine of the participle).

I would have expected ἄδρυνσις to have been ἄδρῡσις, as the neo-Platonist Simplicius (in the commentary on Epictetus’ Encheiridion p. 32) would have spelt it in the 6th century CE, but Aristotle (Metaphysica 1065b20) spelt it with a nu in the 4th century BCE. Why is that? Buck (1933; 151) for example says “Att. ὕφανσις (ὑφαίνω) etc., were formed later and retained ν”.

The historical, or maybe theoretical, formation of the word seems to explain why it has not lost its nu as well. A quick search on the Thesaurus Linguae Graecae shows that, while the cluster -νσ- is reasonably well attested in the whole period of ancient (Attic) Greek (until 1453 CE), the greater part of the attestations are Latin loanwords, foreign place names, or compounds starting in ἐν-, παν-, or συν- that have restored their nu (cf. Buck 1933; 151 ad 3a). Besides words similar to the construction of ἅδρυνσις, the cluster is otherwise very rare in the corpus of the TGL.

In the following two blogs I will delve more deeply into this cluster, and the type of words related to ἅδρυνσις. Next time I will try to make sense out of other words with the cluster -νσ-. In the final blog, we will trace the attestations of ‘real’ Greek -νσ-clusters down to the sources.

  1. Compare Helmut Rix’s Historische Grammatik des Griechischen (1992, pp. 77-79 and 89-91) for all exceptions to these general rules. (Available on Google Books

  2. Masculine παντ-ς > παν-ς > πᾶς loses the nu because it is a final cluster. Neuter παντ > πάν loses the final tau because Greek does not permit final clusters. 

  3. The development of *ti and *ti̯ are essentially the same.