While allowing for polysemy, scholars seem mostly averse to ambiguity, as in the πίστις Χριστοῦ debate; but, it would seem, without engaging with ancient semantic theory. There the model of ‘naming’ and so of evoking an otherwise unspecified mental impression, predominates. Meaning is taken to lie in the mind, not in the word or words that are hoped to evoke it, as is also shown in ancient discussions of metaphor, allegory, and paraphrase. Connotations of individual words are rarely distinguished, rarely if ever purged. We are not justified in expecting verbal precision where our ancient authors will neither have attempted it nor will their hearers have expected it; nor, indeed, do modern psycholinguists appear to find space for it.