Phonology and Phonetics
Voice-Place-Manner (VPM) — Classify consonants systematically: first ask "voiced or voiceless?", then "where in mouth?" (bilabial→velar), finally "how?" (stop, fricative, nasal). Speeds IPA chart navigation and transcription tasks.
Minimal-pair spotting — To find phonemic contrasts, swap one sound in a word and check meaning change (e.g., *pat/bat*). If meaning differs, the sounds are separate phonemes; if not, they're allophones.
Stress-shifts = compound test — In English-type languages, noun compounds often front-stress (*BLACKboard*), phrases stress second (*black BOARD*). Use this pattern to disambiguate morphology in unfamiliar data.
Tone diacritic clock — IPA tone marks: high = top ( ̋ ), mid-high = upper-right ( ́ ), mid = middle ( ̄ ), mid-low = lower-right ( ̀ ), low = bottom ( ̏ ). Visualize a clock face to remember vertical positioning.
Morphology
Case-color code — Memorize core cases: NOM=subject, ACC=object, GEN=possessor, DAT=recipient, LOC=location, INS=tool. In tables, mentally color-code columns so you track which slot does what across paradigms.
Agglutination = Lego blocks — Each morpheme is a separate "brick" with one meaning (Turkish: *ev-ler-im-de* = house-PL-my-in). If you can split every piece cleanly, it's agglutinative; if forms melt together, it's fusional.
Kinship = ego anchor — Always mark "ego" (reference person) first. Then draw: generation (up=parent, down=child), lineage (mother's vs. father's side), and sex. Four parameters solve 90% of kinship puzzles.
Verb-slot template — Many languages order verb morphemes rigidly (e.g., person–tense–mood–aspect). Write out the slot sequence as a formula; fill blanks left-to-right to predict any verb form.
Syntax
SOV vs. SVO fingerprint — SOV languages often put postpositions, adjectives before nouns, genitives before nouns; SVO reverses these. Spot two correlations to guess the rest (Greenberg's universals).
Ergativity triangle — Ergative: S=O≠A (intransitive subject patterns with transitive object). Accusative: S=A≠O. Draw a triangle (S, A, O) and mark which pair gets the same case.
Question-word immobility — Some languages keep question words *in situ* (e.g., Mandarin "you eat what?"), others front them ("What do you eat?"). Check word order first to avoid mis-parsing.
Relative-clause gap — Find the "missing" noun in the relative clause (the gap). English: "the book [I read __]"—gap = object. Identifying gap position clarifies which noun is being modified.
Semantics and Pragmatics
Deixis = speaker's zero point — "Here," "now," "I" only make sense from the speaker's position. In translation, re-anchor these to the new speaker/time/place context; don't translate literally.
Compositionality test — Meaning = parts + structure. If "greenhouse" = green + house (compositional) it's a house that's green; if idiomatic, it's a glass plant-building. Break down each word to check.
Entailment arrow — A entails B if A→B always (no exceptions). "Run" entails "move"; "cat" doesn't entail "black." Draw arrows to map lexical hierarchies (hypernyms/hyponyms).
Numeral Systems
Base detection formula — If numbers repeat/recycle every *n*, base = *n*. E.g., "ten-one, ten-two…" after 10 → base-10; "twenty-one, twenty-two…" after 20 → base-20 (vigesimal).
Body-part counting clue — Vigesimal (base-20) often comes from fingers + toes; base-5 from one hand; base-10 from both hands. Check for "hand," "foot," "person" roots in numerals.
Subtraction trap — Some systems say "two from ten" for 8 (Latin *duodeviginti* = 18). Look for "from," "lacking," or small number *before* large to catch subtractive logic.
Writing Systems and Transliteration
Abugida = consonant king — In abugidas (Devanagari, Ethiopic), consonant is the base letter; vowels are diacritics. Default vowel is /a/; other vowels modify the glyph (e.g., क /ka/, कि /ki/).
Syllabary grid — Japanese kana: rows = consonants, columns = vowels. Fill a 5×5 grid (ka, ki, ku, ke, ko…) to visualize the entire inventory and predict missing syllables.
Logographic = morpheme per glyph — Chinese characters map to morphemes (meanings), not sounds. One character = one syllable = one meaning unit. Count morphemes, not letters, when segmenting.
Historical and Comparative Linguistics
Regular = exceptionless — Sound changes apply to *all* instances in the environment (e.g., Latin /k/→French /ʃ/ before /e,i/: *centum*→*cent*). One counterexample breaks the rule; always check every word.
Cognate checklist — True cognates: (1) similar form, (2) same/related meaning, (3) regular sound correspondence, (4) not borrowed recently. All four required; watch for false friends (e.g., English "much" ≠ Spanish "mucho," borrowed).
Computational and Formal Linguistics
Bracketing for ambiguity — Sentence ambiguous? Write two parse trees with different bracketing: [old [men and women]] vs. [[old men] and women]. Visual structure reveals the two readings instantly.
Rule = if-then algorithm — Formal rules: "Rewrite X as Y in context Z." Write X→Y / Z__. Check left context, target, right context in that order when applying rules to data.