The lexicographer's dilemma

This week's readings presented a dynamic view of the relationship between syntax and semantics. Together, Atkins, Levi, Haspelmath and Slobin put forth a nuanced picture of what it means to be something as simple as a word. They have also added dimension to my conception of "base" words (lexemes), and have increased my appreciation for the dilemma that dictionary makers.

Levin and Atkins explore the use of near-synonyms and highlight from their research several interesting facts about the relationship of word meaning and syntax. The core point of their paper was that general meaning does not determine the way a word is used. Instead, the relationship of semantics and syntax is much more subtle. In the case of the near synonyms explored, senses of internal verses external causation, more than general word meaning, determined how a word was used in a sentence (e.g. when "quiver" is used in an "internal causation" sense, it appears as a transitive verb syntactically). From Haspelmath, we learn about word-forms and lexemes, and that lexicographers have taken to list words with the same inflection all under the same root, where as word of the same family but of different inflection as separate entries in their logs. From there, we complicate our understanding of words by defining "morphemes." This complicates our understanding of what it means for a set of letters to hold meaning (a set of letters need not be a full "word" to hold meaning), especially when the "meaning" of the morpheme can be abstract, or when morphemes themselves just sound different (two different sounds for making nouns plural in English).  And even as Slovin explores the connection of verb phrases in crafting narrative, we see another layer that effects the way we use language (like manner-salience).

I studied Ancient Greek in high school, and was forced to learn six different "principle parts" of verbs, all of different tense, voice and mood. The reason for memorizing six different forms was that a single form did not give you enough information- you could not always accurately predict the form of the 3rd person perfect passive with only the first person present active.

A lexicographer, then, though understanding that "logic" and "logician" are clearly related (and that the meaning of the term can be inferred) must still add it as an entry to a dictionary- not only for the purpose of defining meaning, but also to explain that "logicist" is not a proper English word. Like principle parts in Greek, the form of the word meaning "to practice or study logic" can not be accurately predicted.

Levin and Atkins' study of near-synonyms complicated this view of base words even further. In exploring the concept of lexemes, I realized that the choice of principle parts, as the choice of base words, is arbitrary. Though logic and logician cannot be implied, and so both need to be learned or memorized by language learners, words of the same base with different inflection could be substituted for each other. If dictionaries instead used the participle form of a verb (running, loving) instead of the imperative or 3rd person present (run, love) we would still be able to infer the other inflections of the word. Why was the form of the word that was chosen chosen? Convenience?