Comparative method Study Guide

📖 Core Concepts Comparative Method – Systematic comparison of two or more attested languages to uncover regular phonological & semantic correspondences and reconstruct a proto‑language (the hypothesized common ancestor). Genetic Relationship – Established when many regular, systematic correspondences exist that cannot be explained by universal tendencies or borrowing. Proto‑Language – The reconstructed ancestor language; its forms are proto‑phonemes (e.g., \m, \p) inferred from the data. Regular Sound‑Change Principle – Sound changes are rule‑governed, exception‑free (Neogrammarian hypothesis). Subgrouping – Groups of languages that share innovations (new features) not found in the parent language; innovations, not retentions, signal closer relationships. Parsimony (Occam’s Razor) – Prefer the reconstruction that explains the data with the fewest independent changes. --- 📌 Must Remember Criteria for genetic link: numerous, regular, systematic correspondences & no plausible borrowing or universal explanation. Proof of relationship: successful reconstruction of cognate sets that are semantically aligned. Borrowing ≠ relatedness: Heavy lexical loans (e.g., Persian ← Arabic) do not create a genetic link. Shared innovations > shared retentions for defining subgroups. Neogrammarian claim: sound laws have no exceptions; irregularities are due to analogy, diffusion, or borrowing, not the law itself. Incompatible methods: glottochronology (assumes constant lexical replacement) and mass lexical comparison (lacks sound‑law verification). --- 🔄 Key Processes Assemble potential cognate list – Focus on basic vocabulary (kinship, numbers, body parts, pronouns). Identify and discard borrowings – Use historical knowledge; e.g., English taboo ← Tongan is a loan, not a cognate. Establish correspondence sets – Align sounds across languages (e.g., Hawaiian k ↔ other Polynesian t). Check regularity – Correspondences must recur in many sets; isolated matches are treated as chance. Analyze complementary distribution – Determine if different surface sounds reflect a single proto‑phoneme conditioned by environment (e.g., Verner’s Law). Reconstruct proto‑phonemes – Apply parsimony and typological plausibility (choose \m over \b if the former requires fewer changes). Validate typological consistency – Ensure the reconstructed inventory respects known phonological constraints (symmetrical stop/nasal systems, etc.). --- 🔍 Key Comparisons Borrowing vs. Genetic Relatedness Borrowing: lexical item enters a language from another without systematic sound correspondences. Genetic: systematic correspondences across many basic items, plus shared innovations. Retentions vs. Innovations Retentions: features inherited unchanged from the parent; do not define subgroups. Innovations: new changes unique to a branch; define subgroups. Tree Model vs. Wave Model Tree: clean branching, each node (proto‑language) is distinct and non‑overlapping. Wave: overlapping isoglosses; changes spread outward, can blur strict branching. --- ⚠️ Common Misunderstandings “All similar words are cognates.” – Many lookalikes are loans or chance resemblances. “A single sound correspondence proves relationship.” – Need multiple regular correspondences. “Proto‑languages are fully known.” – Reconstructions are hypotheses; they may be revised with new data. “Tree models perfectly reflect reality.” – Post‑divergence contact (Sprachbunds, diffusion) can violate a pure tree. --- 🧠 Mental Models / Intuition “Correspondence → Proto‑phoneme” – Imagine each language as a transparent sheet; aligned letters across sheets point back to a single underlying symbol. “Innovation as a fingerprint” – A subgroup’s unique change is like a fingerprint that only members share; retentions are generic background patterns. “Parsimony as budgeting” – Each hypothesized change costs a “dollar”; the best reconstruction spends the least total. --- 🚩 Exceptions & Edge Cases Lexical diffusion – A sound change spreads gradually; early adopters may retain the old form, creating apparent irregularities. Analogy – Internal restructuring can make a word resemble another (e.g., Russian devyat’ “nine” reshaped by pattern of sed’ “seven”). Non‑inherited features – Elements lost in all daughter languages (e.g., Latin case system in Romance) cannot be reconstructed. --- 📍 When to Use Which Comparative Method – When you have attested languages with enough basic vocabulary to establish regular correspondences. Lexicostatistics / Swadesh list – Useful for a quick, rough estimate of lexical similarity, but not for proving genetic relationship. Tree model – Apply when innovations clearly delineate clean splits and contact is minimal. Wave model – Prefer when overlapping areal features (Sprachbund) are evident. --- 👀 Patterns to Recognize Systematic sound shifts (e.g., Grimm’s Law: PIE p → f in Germanic). Conditioned environments – Changes that only happen next to certain sounds (Verner’s Law: voicing when the PIE accent is not on the preceding syllable). Complementary distribution – Two surface sounds never appear in the same phonological context → likely allophones of one proto‑phoneme. Cluster of shared innovations – A set of languages showing the same novel sound change or grammatical marker usually form a subgroup. --- 🗂️ Exam Traps Distractor: “Borrowed word = cognate.” – Exams may list a loanword as a cognate; check historical directionality. Trap: “Single correspondence proves relationship.” – Look for multiple, regular sets; a lone match is often chance. Misleading choice: “Glottochronology is a reliable dating method.” – Remember it’s incompatible with the comparative method. Confusing retention with innovation – An answer that cites a shared old feature as evidence of subgrouping is wrong; only new shared features count. Over‑reliance on the tree model – Questions may highlight areal diffusion; selecting a strict tree explanation would be a trap.