Parsing: A Reliable Step-by-Step Method

Parsing means identifying a form’s grammatical features and explaining its job in a sentence. Translation comes afterward. A dependable parser moves from visible evidence—stem, ending, agreement, and clause structure—to meaning, instead of reading left to right as though Latin used English word order.

The method below is deliberately repetitive. At first it feels slower than guessing; soon it becomes the quickest way to avoid guesses. It works for simple dialogue and for dense Classical prose, although literary ellipsis, poetic word order, and archaic forms require a larger inventory of possibilities.

The seven-pass method

The numbering begins with copying because a missed macron, -que, or -ne can derail every later step. In editions without macrons, note quantity-based alternatives and let syntax and context decide.

Pass 0: preserve the evidence

Copy the form exactly. Separate enclitics only after recognizing them: senātus populusque contains populus + -que; venitne? contains venit + -ne. Check whether an editor marks vowel quantity, because quantity can distinguish forms that ordinary ancient and modern editions spell alike.

Present legit has short e; perfect lēgit has long ē. An unmacronized text prints both as legit. The adverbs cotīdiē and herī then become essential evidence, but they confirm a morphological possibility rather than manufacture a tense from nothing.

Pass 1: find finite-verb anchors

Finite verbs carry person and number and therefore define clause centers. Circle them before assigning noun functions. Imperatives also anchor clauses even when their subject “you” is unexpressed.

Mitte is a second-person singular present active imperative of mittō; its subject is an understood tū. Est is third-person singular present indicative and agrees with nominative māter. The semicolon separates the two clause centers.

Ending -istī identifies second-person singular perfect active. Perfect stem ēg- points to agō, agere, ēgī, āctum, not to a hypothetical verb beginning ēg-. Interrogative quid is the direct object.

Pass 2: list morphological possibilities

For each noun or adjective, state case, number, and gender; do the same for a pronoun wherever those features are distinguished. For each finite verb, state person, number, tense, voice, and mood. Do not assign a syntactic function merely because one parse resembles English order.

Puellae can be nominative plural, genitive singular, or dative singular; vocative plural is also formally possible. Rosās is accusative plural. Dant is third-person plural present active indicative. With no wider context, nominative plural puellae gives “the girls give roses,” but dative singular gives “they give roses to the girl,” with an unexpressed plural subject. Morphology proves that both readings are possible.

Pass 3: recover the dictionary entry and stem family

Strip only a genuine ending, then account for any stem alternation with the dictionary entry. Noun genitives and verb principal parts are decisive. If the remaining letters do not match a known stem, reconsider both the ending and the lemma.

Tulērunt is third-person plural perfect active: perfect stem tul- plus -ērunt. The entry is ferō, ferre, tulī, lātum. No manipulation of present fer- alone predicts tul-; the alternate stem is lexical information.

Accusative plural clāvēs points to clāvis, clāvis f.; meās agrees in case, number, and gender. First-person singular inveniō supplies the unexpressed subject “I.” Parsing keeps clāvēs from being mistaken for a nominative merely because it comes first.

Pass 4: connect agreement and government

Agreement creates visible networks. An adjective or participle normally matches its noun in case, number, and gender; a finite verb normally matches its subject in person and number. Government creates another network: prepositions require cases, and certain verbs or adjectives require particular complements.

Mīlitēs is formally nominative or accusative plural. Ducem is accusative singular. Masculine nominative plural secūtī agrees with mīlitēs, and plural sunt completes the perfect of deponent sequor. The agreement network selects mīlitēs as subject; the deponent dictionary entry selects active meaning.

Carrō tuō is ablative singular because deponent ūtor governs the ablative. English use takes a direct object, so English syntax is actively misleading here. Dictionary government outranks transfer from English. The dictionary-entry guide explains how such case requirements are recorded.

Pass 5: build constructions and clauses

After local connections are secure, identify larger constructions: infinitive clauses, participial phrases, ablative absolutes, and subordinate clauses introduced by words such as cum, ut, or sī. Mood is often a structural clue.

Hōc, librō, and lēctō are all ablative singular. The noun and participle form an ablative absolute, grammatically detached from the implied subject of discessit. A literal word string—“this book read, he left”—records forms but not syntax.

Pluperfect subjunctive vēnisset belongs to the cum clause; pluperfect indicative clausae erant anchors the main clause. Feminine nominative plural clausae agrees with portae. The opening portae could have other parses in isolation, but agreement and the plural verb resolve it.

Dīxit is the finite anchor. Accusative sē is the subject of infinitive esse, and future active participle ventūrum agrees with it. This accusative-and-infinitive construction is normal Classical reported speech; English inserts “that” and a finite verb.

Passes 6 and 7: choose the function, then translate

Once morphology and syntax have reduced the possibilities, use discourse and real-world context. Only then produce natural English. A structurally faithful analysis may require a nonliteral final translation.

The Latin structure is mihi “to me” + nominative liber + est: possession with sum and a dative possessor. “A book is to me” is useful as a temporary structural gloss, not acceptable final English. In the relative clause, quem is accusative object of dabō, while dative tibi is the recipient.

A compact parsing template

For a nominal form, report: part of speech; case, number, gender where applicable; lemma; agreement; function. For example: “secūtī: perfect participle of deponent sequor; nominative masculine plural; agrees with mīlitēs; with sunt forms the perfect.”

For a finite verb, report: person, number, tense, voice, mood; principal parts; subject; clause function. For example: “tulērunt: third-person plural perfect active indicative of ferō, ferre, tulī, lātum; implied subject ‘they’; main-clause predicate.”

In verse, poets may separate agreeing words widely or use contracted and alternative forms (literary). Early inscriptions may preserve endings absent from standard Classical prose (archaic). The workflow does not change: expand the formal candidate list using the appropriate register, then test agreement and syntax.

Common Mistakes

1. Making the first noun the subject

2. Choosing the first familiar meaning of an ending

3. Guessing the lemma from the visible beginning

4. Stopping at a word-for-word gloss

5. Naming forms without connecting agreement and syntax

Key Takeaways

Copy accurately, find finite anchors, list every viable morphology, recover the lemma, connect agreement and government, build clause structure, resolve ambiguity from context, and translate last. The central habit is simple: let Latin forms establish the relationships before English word order suggests a story.