1.10 English Irregularities and Cross-Linguistic Comparison
Key Takeaways
- English irregular verbs are historical survivals of Germanic strong-verb vowel gradation, not random exceptions.
- Irregular plurals fall into predictable families: mutation plurals, zero plurals, -en plurals, and retained Latin/Greek forms.
- Overgeneralization errors such as 'goed' and 'childs' signal that the learner has acquired the regular rule and is applying it productively.
- Cognates give Spanish-speaking learners access to thousands of academic words, but false cognates such as 'embarazada' and 'constipado' create predictable errors.
- Languages differ systematically in phonology, morphology, syntax, semantics, and pragmatics, and those differences predict which English features a given L1 group will find hardest.
1.10 English Irregularities and Cross-Linguistic Comparison
Two ETS sub-topics converge here. The first states that the candidate "knows about the inconsistencies and irregularities of the English language" and "identifies irregular verbs, irregular nouns (plurals), and irregular spelling." The second states that the candidate "knows that languages differ from or are similar to each other in their phonology, morphology, syntax, semantics, and pragmatics" and "identifies cognates." They belong together because irregularity is exactly where learners fall back on their first language — and where cross-linguistic comparison predicts what will happen next.
1. Irregular Verbs Are a System, Not a List
English has roughly 200 irregular verbs, but they include the highest-frequency verbs in the language — be, have, do, go, say, make, take, come, see, get. A learner cannot avoid them. They are the residue of the Germanic strong verb system, which marked tense by changing the stem vowel (ablaut) rather than by adding a dental suffix.
| Pattern | Base → Past → Past participle | Examples |
|---|---|---|
| Invariant (all three same) | cut → cut → cut | cut, put, hit, cost, let, set, shut, hurt |
| Two-part (past = participle) | buy → bought → bought | bring/brought, teach/taught, catch/caught, think/thought |
| Vowel gradation, three distinct forms | sing → sang → sung | drink, swim, ring, begin, shrink |
| -ow/-ew/-own | know → knew → known | grow, throw, blow, fly (flew/flown) |
| -eak/-oke/-oken | speak → spoke → spoken | break, choose (chose/chosen), freeze |
| Suppletive (unrelated stems) | go → went → gone | be (am/is/was/were/been) |
Teaching these by pattern family rather than as an alphabetical list gives learners a generative hook: once sing/sang/sung is secure, ring, drink, and swim follow. This is a direct application of the morphological-analysis principle in Section 4.2.
2. Irregular Plurals
| Family | Rule | Examples |
|---|---|---|
| Mutation (umlaut) plurals | Internal vowel change | man/men, woman/women, foot/feet, tooth/teeth, goose/geese, mouse/mice, louse/lice |
| -en plurals | Old English weak plural | child/children, ox/oxen |
| Zero plurals | No change | sheep, deer, fish, series, species, aircraft, salmon |
| -f → -ves | Voicing of the final fricative | leaf/leaves, knife/knives, wife/wives, life/lives (but roof/roofs, chief/chiefs) |
| Latin/Greek retained | Source-language morphology | criterion/criteria, phenomenon/phenomena, analysis/analyses, datum/data, index/indices, cactus/cacti |
The Latin and Greek plurals matter disproportionately in secondary content classes: criteria, data, analyses, hypotheses, and phenomena are Tier 2 and Tier 3 academic vocabulary (Section 4.2), and a student who writes "the datas shows" has an academic-register problem, not a basic grammar problem.
3. Overgeneralization Is Evidence of Learning
When a learner writes "The two childs bringed their foots," the diagnosis is overgeneralization of a productive regular rule — an intralingual developmental error. The learner has acquired add -s for plural and add -ed for past and is applying them systematically. This is the same U-shaped developmental curve seen in first-language acquisition, where children produce went, then goed, then went again.
The instructional consequence: do not treat overgeneralization as regression. It signals rule acquisition. Respond with high-frequency irregular form practice in meaningful contexts, not with a re-explanation of the regular rule the student has clearly already learned.
4. Irregular Spelling
English spelling irregularity has identifiable historical causes (Section 1.6): the Great Vowel Shift left spellings frozen after pronunciation moved (name, time), the printing press fixed conventions during a period of change, and heavy borrowing imported source-language orthography (ballet, psychology, bologna). The practical categories:
- Silent letters: knee, write, lamb, island, debt, honest, Wednesday
- One sound, many spellings: /iː/ as ee, ea, ie, ei, e, y — see, sea, chief, receive, be, city
- One spelling, many sounds: -ough in through, though, thought, tough, cough, bough
- Homophones: their/there/they're, to/too/two, write/right/rite
5. Cross-Linguistic Comparison: Predicting Difficulty
Languages differ systematically at every level, and those differences predict learner difficulty. This is the contrastive framework introduced in Section 2.3, applied concretely.
| Level | Typical English feature | Contrasting L1 pattern | Predicted learner difficulty |
|---|---|---|---|
| Phonology | /r/–/l/ contrast; final consonant clusters | Japanese has one liquid; Spanish disallows /sp-/ onsets | right/light confusion; especial for special |
| Morphology | Analytic, few inflections | Turkish, Korean are agglutinative; Arabic is root-and-pattern | Over- or under-marking; word-boundary uncertainty |
| Syntax | Rigid SVO; obligatory subjects | Japanese/Korean SOV; Spanish/Italian pro-drop | "The student the homework completed"; "Is very important" |
| Semantics | make/do, say/tell distinctions | One verb covers both in many L1s | "do a mistake," "say me the answer" |
| Pragmatics | Indirect requests signal politeness | Directness is neutral or polite in many cultures | Requests heard as rude (Section 1.4) |
| Orthography | Latin alphabet, deep orthography | Arabic (right-to-left, abjad); Chinese (logographic) | Directionality, letter formation, vowel representation |
Cognates: the highest-leverage transfer resource
Roughly 30 to 40 percent of English words have a Spanish cognate, and the proportion rises in academic and technical vocabulary because both languages draw on Latin and Greek roots: analyze/analizar, observe/observar, photosynthesis/fotosíntesis, democracy/democracia. Explicit cognate awareness instruction — teaching students to look for the resemblance and to check whether the meaning fits — measurably accelerates academic vocabulary growth for Spanish-speaking learners.
False cognates (false friends) are the countervailing trap:
| Spanish word | Looks like | Actually means |
|---|---|---|
| embarazada | embarrassed | pregnant |
| constipado | constipated | having a head cold |
| éxito | exit | success |
| librería | library | bookstore |
| asistir | assist | to attend |
| actualmente | actually | currently |
| sensible | sensible | sensitive |
| carpeta | carpet | folder |
The pedagogically correct move is to teach cognate strategy and the verification habit together: notice the resemblance, predict the meaning, then confirm it against context. Teaching only the trap suppresses a powerful strategy; teaching only the strategy produces confident errors.
A second-grade multilingual learner who previously said 'went' correctly now says 'goed' and 'bringed.' What does this shift indicate?
A Spanish-speaking high school student writes in a health class journal: 'My mother is embarrassed and the baby arrives in June.' What is the most likely source of this error?
Which set of English plurals is formed by internal vowel change (mutation) rather than by suffixation?
An ESOL teacher wants to accelerate academic vocabulary growth for a class of Spanish-speaking newcomers studying biology. Which approach best reflects the research on cognates?