Common Phonological Rules

Why This Matters

Phonological rules are the hidden machinery behind every word you speak. When you're tested on phonology, you're not just being asked to identify what happens to sounds—you're being asked to explain why sounds change in predictable ways. These rules demonstrate core linguistic principles: economy of effort, perceptual clarity, feature spreading, and systematic sound change. Understanding these processes connects directly to bigger course concepts like language acquisition, historical linguistics, and dialectal variation.

Here's the key insight: phonological rules aren't random. They follow patterns based on articulatory ease, acoustic distinctiveness, and the phonological features of surrounding sounds. Don't just memorize rule names—know what phonetic motivation drives each process and be ready to identify examples in transcription data. That's what separates a strong exam answer from a mediocre one.

---

Processes That Make Sounds More Similar

These rules operate on a simple principle: sounds influence their neighbors to share features, reducing the articulatory effort required to move between them.

Assimilation

• Sounds become more like neighboring sounds—this can affect place of articulation, manner, or voicing features
• Nasal place assimilation is a classic example: "input" surfaces as [ˈɪmpʊt] because $n$ adopts the bilabial place of the following $p$
• Most common in connected speech, where the pressure for articulatory efficiency is highest—this is why casual speech sounds different from careful citation forms

Vowel Harmony

• Vowels within a word share specific features—typically backness, rounding, or ATR (advanced tongue root)
• Agglutinative languages like Turkish and Hungarian show this dramatically: suffixes change their vowels to match the root (ev-ler "houses" vs. köy-ler "villages")
• Demonstrates feature spreading across multiple syllables—a key concept for understanding autosegmental phonology

Palatalization

• Consonants shift toward the palate when adjacent to front vowels or glides like $j$
• Place of articulation changes—velars often become palatals or affricates (Latin $k$ before $e/i$ became Italian [tʃ])
• Extremely productive cross-linguistically, making it a go-to example for sound change questions on exams

---

Processes That Make Sounds Less Similar

Sometimes similarity creates problems. When two sounds are too alike, languages may change one to improve perceptual distinctiveness or ease articulation.

Dissimilation

• Similar sounds become less alike—the opposite of assimilation, driven by the need to avoid confusing repetition
• The OCP (Obligatory Contour Principle) often motivates dissimilation: languages disprefer identical adjacent elements
• Historical examples include Latin peregrinus becoming Spanish peregrino then English "pilgrim" ($r...r$ → $l...r$)

---

Processes That Add or Remove Sounds

These rules change the number of segments in a word. The motivation is typically to simplify difficult sequences or to break up clusters that violate a language's phonotactic constraints.

Insertion (Epenthesis)

• A sound is added to break up consonant clusters or satisfy syllable structure requirements
• Vowel epenthesis is common: "athlete" → [ˈæθəlit] inserts a schwa to avoid the [θl] cluster
• Consonant epenthesis also occurs: English "warmth" historically added [p] between [m] and [θ]

Deletion (Elision)

• A sound is removed, typically in unstressed positions or rapid speech contexts
• Syncope (medial deletion) and apocope (final deletion) are specific subtypes worth knowing
• "Next day" → [nɛks deɪ]—the $t$ deletes because three consonants in a row violates English phonotactics

Metathesis

• Sounds switch positions within a word, often to create more preferred syllable structures
• Common in child language ("spaghetti" → "pasketti") and historical change ("bird" was Old English brid)
• Can become lexicalized—"ask" pronounced as [æks] in some dialects reflects a metathesis that has stabilized

---

Processes That Change Sound Strength

Consonants exist on a sonority scale from most obstructed (stops) to most open (vowels). These rules move sounds along that scale, either weakening or strengthening them.

Lenition

• Sounds become weaker—stops may become fricatives, fricatives may become approximants, or sounds may delete entirely
• Intervocalic lenition is especially common: Latin $p, t, k$ between vowels became Spanish $b, d, g$ (voiced) or disappeared
• Represents a trajectory: stop → fricative → approximant → deletion—know this scale for analyzing sound changes

Fortition

• Sounds become stronger—the reverse of lenition, often involving devoicing or stop formation
• Word-initial position frequently triggers fortition: approximants may become fricatives or stops at utterance beginnings
• Less common than lenition overall, but important for understanding phonological asymmetries

Final Devoicing

• Voiced obstruents become voiceless at word boundaries—a specific type of fortition/neutralization
• German and Dutch are textbook examples: German Rad "wheel" and Rat "advice" both end in [t]
• Creates neutralization—underlying contrasts disappear in certain positions, a key concept for phonemic analysis

---

Quick Reference Table

---

Self-Check Questions

1. Both assimilation and vowel harmony involve feature spreading—what distinguishes the domain over which each process operates?

2. You observe that a language changes $t$ to [s] before $i$. Which phonological rule is this, and what phonetic motivation explains it?

3. Compare epenthesis and deletion: if a language has strict CV syllable structure and encounters a word-final consonant cluster, which process would you predict, and why?

4. A student transcribes German Hund "dog" as [hʊnd] but a native speaker produces [hʊnt]. What rule accounts for this, and what does it tell us about underlying vs. surface forms?

5. Lenition and fortition are opposite processes, yet lenition is far more common cross-linguistically. What does this asymmetry suggest about the directionality of sound change?