Skip to main content

In my paper I examine the use of analogy as a lingusitic principle in Pollux' lexicon (Onomastikon) of the 2nd century CE.

Pollux' Onomasticon is mainly a lexicon of synonyms, derivatives and antonyms, organised according to semantic fields. In that respect it continues in the Hellenistic tradition of descriptive, ideographic dictionaries, although unlike them it doesn't focus on the vocabulary of only one subject but covers diverse areas. At the same time, the Onomasticon shares a number of key features with the prescriptive Atticistic lexica: it systematically characterizes certain words as "bad" or "unacceptable" and it declares the first and most important criterion for a word’s inclusion to be its attestation in one of the canonical Attic authors. The principle use of the Onomastikon was as an aid to orators in their selection of appropriate words for composion and in delivering speeches.

These two contradictory starting points create a tension which informs the Onomastikon throughout and cannot be resolved through an emphasis on any one of the two (see e.g. Tosi 2013 for emphasis on the descriptive tendencies and Matthaios 2013 who accentuates the perscriptive/atticistic character of the lexicon).

I will argue that an awareness of this tension is indispensable for our understanding of the Onomastikon. My focus here will be on examples where Pollux uses arguments through analogy. On a number of occasions a recommended (or at least tolerable) word is not supported by reference to an ancient author, but through the development of an argument. Most of these arguments are based on analogy and are employed in the defence of a word, or a certain semantic use of it. Following on from a typology of the various types of argumentation, I will focus on five characteristic examples: μετοπωρίσαι (1.62), νακοτίλτης (7.28), σάρον (10.28--9), πινάκια (10.82–4) and παροψίς (10.87–8). My aim is to show how the Atticistic principle of approving words only attested in canonical authors, although not totally abandoned, is circumvented by analogical argumentation. This enables Pollux to “defend” words or meanings that have found their way into the educated language of his time.

Besides cases of overt argumentation, there are many examples of words or meanings unattested in Attic which are nonetheless without any argumentation comprised in the Onomastikon. It can plausibly be shown that some of these are “accepted” on the principle of analogy. I will discuss the examples of ἀβέβηλος (1.8; a synonym of ἱερός and an antonym of βέβηλος referring to a place dedicated to a god) and ἀργυρίτης (3.153; antonym of στεφανίτης referring to a sport competition). It is worth noting that the adjective ἀβέβηλος is followed by the comment "although I didn't ow meet the word anywhere up until now "; “anywhere” means in conformity to the atticistic praxis “anywhere in the canonical attic texts”. The creation of a complete pair of antonyms βέβηλος – ἀβέβηλος in analogy to a general pattern of the Greek language, while simultaneously "defending" a word that is part of educated vocabulary (e.g. Josephus BJ 4.242, Plu. Mor. 166 E 11), is for Pollux much more important than the fact that the word is not attested in the canonical authors.

Pollux' use of analogic argumentation and of analogy has an important implication for the placement of the Onomastikon in the history of ancient linguistics. It is a feature that connects Pollux to the major current in Hellenistic linguistics, which is based on the postulation of analogy in language. During Pollux lifetime this current becomes less dominant and analogy is rather insignificant in other Atticistic lexica. Pollux is in contact with this tradition while at the same time adopting the programme of Atticism.