Internal reconstruction is a method for recovering earlier stages of a language using evidence from within a single language, without requiring comparison to related languages. It exploits synchronic alternation patterns — cases where the same morpheme surfaces in different phonological shapes depending on its environment — to infer the historical changes that produced them. For example, if a root alternates between [k] and [tʃ] before different vowels, internal reconstruction posits an earlier uniform form that was split by a conditioned sound change. Analogical leveling, where paradigm irregularities are eliminated by extending the pattern of one form to others, can obscure the evidence that internal reconstruction depends on, making alternation patterns progressively less transparent over time. The method is particularly valuable for language isolates, where no relatives exist for comparative work.
Examine morphophonemic alternations in a familiar language (English "electric/electricity," "divine/divinity") and reason backward to the earlier state that produced them. Work through a problem set where you are given a paradigm with allomorphic variation and must reconstruct the pre-alternation form plus the sound change responsible. Compare your internal reconstruction with the known historical record (e.g., Old English or Latin) to check your reasoning.
In your study of sound change and reconstruction, you learned how linguists use the comparative method: gather cognates from related languages, identify systematic correspondences, and reconstruct the proto-language from which they descended. But what do you do when no related languages are available — when a language is an isolate, its relatives long extinct or unknown? Internal reconstruction is the method that turns the question inward: it uses patterns within a single language to infer that language's own history.
The raw material is synchronic alternation — cases where a single morpheme surfaces in phonologically different shapes depending on its environment. In English, consider "electric/electricity": the root-final consonant alternates between [k] and [s] before different suffixes. This alternation is not random; it tracks the phonological environment. Internal reconstruction asks: what was the single earlier form that a conditioned sound change split into these two variants? The answer posits an earlier form with [k] throughout, and a historical palatalization rule that changed [k] to [s] before certain vowels. The alternation visible in present-day English is a synchronic fossil of that historical process.
The method works by reasoning from alternation patterns to the change that must have created them. You posit the simplest earlier form — typically the one appearing in the most basic or most frequent context — and formulate the change that derives the alternating form in the right environment. The conditioning environment is crucial: it reveals both the direction of change (A → B, not B → A) and its phonological trigger. When the same conditioning environment recurs across multiple morpheme pairs showing the same alternation, the reconstruction gains confidence: you are not just explaining one quirk, but a systematic pattern.
A critical limitation must be held clearly in mind: internal reconstruction is vulnerable to analogical leveling. Languages regularize irregular paradigms over time, replacing alternating forms with a single uniform one — eliminating the very surface evidence the method depends on. When leveling has occurred, that chapter of the language's history becomes invisible to the internal method. This is also why internal reconstruction recovers a pre-alternation stage of the language being analyzed, not the proto-language. The comparative method, when available, reaches further back and cross-checks internal reconstruction. Where both methods converge on the same ancestral form, the reconstruction is on solid ground. Where they diverge, there is a research question rather than an answer.