Like Greek and Latin, Coptic does not make word divisions in writing, but linguists have usually assumed that languages nevertheless divide the speech stream into words. This has sometimes been questioned, but mainstream linguistics has shown little interest in marshalling support for the word as a universal unit of language. The distinction between syntax and morphology, which is based on the word, seems too deeply ingrained in our thinking about language structure – it is thus difficult to imagine life without the word, and dizzying to think that it might be nonexistent.
As is well-known, Coptic texts can be divided into stress groups (called "bound groups" by Layton 2004), each of which consists of a noun or verb root, preceded by a series of unstressed, proclitic elements. Calling these elements "prefixes" would result in an extremely strange inflectional system (where nouns can be inflected for tense of the clause, for gender of the relative clause they belong to, or even for predicative action, because some of the proclitics are verbs). Thus, Coptic is a particularly well-suited language to illustrate the fact that "words" are much like "letters" – useful elements for writing, but with no role in linguistic analysis.
Photo: Leaf from a Coptic parchment codex (Berlin, Papyrus collection P 9531) containing parts of the Old Testament Book of Proverbs, chapter 19