Templatic morphology encodes affixation through abstract CV (consonant-vowel) templates rather than linear morpheme sequences. Semitic languages exemplify this: roots (like KTB 'write') are triconsonantal, and vocalic patterns (templates) mark meaning and inflection (kataba 'he wrote', kitaab 'book', ktaab 'books'). This non-linear organization elegantly explains root-pattern morphology and accounts for allomorphy as variation in template application.
Analyze Semitic (or other templatic language) data, identifying root consonants and templates. Show how the same root appears in different forms by varying templates, and explain allomorphic relationships.
From morphological structure, you know that words are built from morphemes — minimal units of meaning — and that morphemes can be arranged as prefixes, suffixes, infixes, or reduplication. All of these are concatenative: morphemes attach sequentially to a base, and you can identify where each one begins and ends. The word *unhappiness* is the sum of *un-* + *happy* + *-ness*, each morpheme a discrete string. Templatic morphology describes a fundamentally different kind of organization: one where morphological information is encoded not as segments added to a string but as a pattern imposed *across* an underlying form.
The clearest examples come from Semitic languages — Arabic, Hebrew, Aramaic, Maltese. Consider the Arabic root KTB, which carries the meaning-core 'write'. This root consists of three consonants but no vowels. To generate actual words, you combine this consonantal root with a vocalism template — a pattern of vowels and sometimes additional consonants that carries grammatical and semantic information. The pattern *CaCaCa* (where C marks a root consonant position) gives *kataba* 'he wrote'. The pattern *CiCaaC* gives *kitaab* 'book'. The pattern *maCCaC* gives *maktab* 'office' or 'desk'. The root is always K-T-B; the different surface forms arise from different templates being applied to the same root. Crucially, you cannot simply segment these words into morpheme + morpheme: the root and the template are interleaved, not concatenated.
The formal representation in autosegmental phonology — which you may have encountered in suprasegmental phonology — uses multiple tiers. The consonantal root lives on one tier, the template (a CV skeleton) on another, and the vowel melody on a third. Association lines connect the tiers: root consonants map to C positions, vowels map to V positions. This is elegant because it captures what is linguistically real: speakers know that *kataba*, *kitaab*, and *maktab* share something (the KTB root), and the multi-tier representation makes that sharing explicit. The same root appears in dozens of words; learning the root and the productive templates lets you generate and recognize a large vocabulary from a small inventory.
The morphosyntactic information carried by templates is considerable. In Arabic, the *CaCaCa* pattern is the perfective third-person masculine singular active verb — one template carries tense, aspect, agreement, and voice simultaneously. The passive counterpart (*CuCiCa*, e.g., *kutiba* 'it was written') uses a different vocalism on the same root. Plural formation in Semitic is similarly templatic: rather than adding a suffix to a singular noun, Arabic often uses a broken plural formed by changing the internal vowel pattern. *Kitaab* (book) → *kutub* (books); *bayt* (house) → *buyuut* (houses). These are not phonologically predictable from the singular by any simple rule — learners must acquire the plural template for each noun class.
The insight that makes templatic morphology theoretically significant is that it separates segmental content (which consonants appear) from prosodic/syllabic structure (how those consonants are arranged in a template). This separation is not available in concatenative morphology, where the two are fused. Recognizing this class of morphological phenomena required generative linguists to move beyond the assumption that morphology is always linear — a move that proved productive not just for Semitic but for understanding autosegmental processes (tone, vowel harmony) across many unrelated languages. Templatic morphology is thus both a description of how Semitic languages work and an argument for a more powerful and flexible theory of what morphological structure can look like.