Verb Stem Classes and Principal Parts: Contrasts and Learner Traps

Hungarian verb alternations become manageable when three questions stay separate: which lexical stem is available, which tense or mood marker follows it, and which person-object ending the clause requires. Similar-looking changes may come from different sources. Olvasson reflects sibilant assimilation, tanítson a t-final imperative pattern, aludjon a stored nonpresent stem, and legyen suppletion.

This page distinguishes those mechanisms through matched pairs. Use the forms page for the principal-part inventory and the usage page for communicative contexts. Unless marked otherwise, forms belong to neutral modern standard Hungarian.

Citation form versus infinitive

Hungarian cites verbs in third-person singular indefinite present. An English dictionary habit—assuming the headword is an infinitive or bare invariant root—therefore fails immediately.

Final -ik marks the citation form alszik, but the infinitive is aludni, not alszikni or alszni. Similarly, megy → menni, tesz → tenni, eszik → enni, and iszik → inni. For regular vár → várni, citation form and infinitive happen to share vár-, but that transparent case is not a universal template.

Stable stem versus vowel loss

Compare a stable verb with a vowel-loss verb in the same person:

Sétál-unk preserves the citation stem. Mozg-unk corresponds to citation mozog and loses medial o in this cell. Yet other present cells retain it: mozogsz, mozog, mozogtok, mozognak.

The shortened pattern is lexical and slot-sensitive:

First-person érzek uses érz-, while third-person plural éreznek uses érez-. Do not invent one short “true stem” and force it everywhere. Forms such as éreztem but third-person érzett show that even the past paradigm can distribute fuller and shorter shapes differently.

Vowel loss versus stem expansion

Both processes create two visible shapes, but in opposite directions. With érez/mozog, the citation form contains a vowel absent from certain inflected forms. With fürdik, the citation form shows fürd-, while many nonpresent and nonfinite forms use fuller füröd-.

The present mainly shows fürd-, while most personal past forms use füröd-, as in fürödtem. Third-person singular is the important exception fürdött, not fürödött. Modern Hungarian permits parallel variants in parts of this paradigm, and fürdeni also occurs beside fürödni. That variation does not make arbitrary mixtures correct: the safest learning strategy is to record the common forms actually used in your target variety and register.

Expanded -szik verbs create a stronger lexical contrast: fekszik–feküdni–feküdt, alszik–aludni–aludt, emlékszik–emlékezni–emlékezett.

The present emléksz- and infinitive emlékez- cannot be derived by one productive sound rule available to new verbs.

Consonant-final regular versus vowel-final v-stem

An ordinary consonant-final verb retains its stem before first-person -ok/-ek/-ök: vár → várok. The small vowel-final class inserts v before certain vowel-initial endings and usually shortens the vowel: lő → lövök, nő → növök, sző → szövök.

First-person lövök has the v-stem; third-person citation lő does not. Before consonant-initial conditional -n-, the long vowel returns:

Do not generalize v insertion to productive vowel-final loans or to nouns. The native verb class is small and closed enough to memorize: fő, lő, nő, nyű, ró, sző and their prefixed derivatives.

Sibilant stem versus t-final stem before j

Sibilants and t both obscure the imperative-subjunctive marker j, but their outputs differ. After s, sz, z, zs, dz, complete assimilation creates a long sibilant: olvas → olvasson, néz → nézzen, edz → eddzen. T-final verbs have several subpatterns.

Olvasson has long ss; tanítson has ts. For a short-vowel-plus-t verb, compare szeressen, fusson. For a consonant-plus-t stem, compare mentsen. High-frequency lát takes exceptional lásson.

Productive alternation versus lexical suppletion

Sibilant assimilation applies to new verbs with the relevant final consonant. Suppletion does not. Van selects unrelated-looking principal parts according to grammatical environment:

Present van, past volt, and imperative-subjunctive legyen are stored members of one lexeme. Conditional lenne and infinitive lenni align with legy- only historically, not by a productive modern replacement rule.

The frequent “seven” set is multistem rather than fully suppletive: tesz–tenni–tett–tenne–tegyen, with parallel vesz, visz, hisz, eszik, iszik, lesz. Their shared shapes help memory, but do not license forms such as mond–monni–mott for regular mond.

Stem alternation versus definite agreement

English speakers often notice a changed verb and attribute every change to irregularity. Compare érez:

Both forms use shortened érz-. -ek marks first-person indefinite with valami szokatlant; -em marks first-person definite with demonstrative ezt. The stem class is constant while object agreement changes.

Prefixes are another independent layer:

Both verbs use irregular past ment. The prefix attaches in the neutral positive clause and follows the verb under negation. Prefix separation does not create a new stem.

Register and competing variants

Some stem variants coexist without a sharp meaning difference. Éneklek/énekelek “I sing” and éneklünk/énekelünk “we sing” both occur in modern usage. Shorter variants are well established, while fuller variants transparently preserve énekel-; this is variation, not an indefinite/definite contrast.

Both are acceptable. In edited prose (formal), consistency within a passage is more useful than inventing a categorical ban. By contrast, alszott is not a stylistic variant of aludt in modern standard Hungarian.

Decision tree

1. What grammatical cell is needed? Identify present/past, conditional/imperative, subject person, and object series.
2. Is the citation stem stable? If regular, attach the harmonic tense/mood/person material.
3. Does the stem end in a sibilant or t? Apply the appropriate j interaction; for t, identify the lexical subtype.
4. Does a vowel-final native verb select a v-stem? Use löv-/növ-/szöv- before the relevant vowel-initial endings, but long lő-/nő-/sző- elsewhere.
5. Does the lexeme have shortened or expanded principal parts? Retrieve érz-/érez-, fürd-/füröd-, alsz-/alud-, or the recorded alternants.
6. Is it suppletive or one of the frequent multistem verbs? Retrieve van–volt–lenne–legyen or tesz–tett–tenne–tegyen directly.
7. Is a prefix or object causing an additional visible difference? Compute prefix order and definite agreement separately.

Diagnostic exercise

Choose the standard form and state what licenses it.

1. “we move”: mozogunk / mozgunk
2. “I would shoot”: lövnék / lőnék
3. “that she should read”: hogy olvason / hogy olvasson
4. “the child slept”: a gyerek alszott / a gyerek aludt
5. “it is important that he be here”: fontos, hogy itt van / legyen
6. “I feel this”: ezt érzek / ezt érzem

Answers: (1) mozgunk uses the vowel-loss stem. (2) lőnék uses long-vowel lő- before conditional -n-. (3) olvasson writes the assimilated long ss. (4) aludt uses the stored past stem. (5) legyen is the suppletive imperative-subjunctive required for the desired state. (6) ezt érzem combines shortened érz- with definite first-person -em selected by demonstrative ezt.

Common Mistakes

Key Takeaways

• Citation form and infinitive coincide transparently only for regular stems.
• Vowel loss, expansion, v insertion, sibilant assimilation, and t-final changes are distinct mechanisms.
• Productive phonology can generalize; lexical principal parts and suppletive forms must be memorized.
• Variant stems such as éneklek/énekelek are not the same phenomenon as nonstandard creations such as alszott.
• Object definiteness and prefix placement add morphology and word order independently of stem class.
• A reliable analysis proceeds from grammatical cell to stem alternant to ending, while respecting Hungarian vowel length and multigraph spelling.