Philological Analysis of Discontinuous Predicates

A Persian lexical predicate can be structurally unified without being typographically or syntactically adjacent. Modern دعوتش کردم places an object clitic between the lexical nonverbal base and the inflected light verb. Classical prose and verse allow heavier intervening material, while manuscripts may supply spacing that reflects scribal habit rather than modern word boundaries.

Philological analysis therefore cannot equate “written together,” “standing next to each other,” and “forming one predicate.” It must combine syntax, argument structure, semantic unity, prosody where recoverable, parallel passages, dictionary evidence, and the edition's own normalization policy.

Four different kinds of separation

The same surface distance can arise from different structures. A clitic is phonologically dependent; a prepositional phrase is a syntactic complement; a modifier may make a nominal element more referential; poetic hyperbaton may displace a constituent for meter or focus. “Discontinuous compound” is thus a conclusion to justify, not a label triggered by any intervening word.

The first sentence invites analysis of سخن کردن “speak” with an expanded nominal domain; the second has the transparent predicate باز کردن. Both end in کرد, but the lexical organization differs.

Start from semantic and argument-structure unity

A useful first diagnostic is whether the pieces describe one event and jointly determine its participants. In دست از کار کشیدن “stop working/give up the task,” literal “hand” is not an independently manipulated object, and از کار is licensed by the idiom as a whole. Contrast literal پرده را کنار کشیدن, where an agent physically pulls a curtain.

Inflection offers another clue. Tense, negation, and agreement occur on the verbal component, but their semantic scope covers the whole predicate: دست نکشید means “did not give up,” not simply “did not pull.” Still, final inflection alone is not sufficient, because ordinary object-plus-verb combinations inflect in the same place.

Classical word order and intervening dependents

Classical Persian was verb-final but permitted discourse-driven rearrangement, heavy constituents, and poetic displacement. A nonverbal predicate element could be expanded or separated from its light verb by complements and modifiers. An editor should ask whether the intervening phrase belongs inside the nonverbal constituent, functions as an argument of the whole predicate, or has been scrambled for focus.

Here پیمانی استوار is internally modified and more noun-like than bare پیمان بستن. That does not force an all-or-nothing decision between “noun plus heavy verb” and “compound verb.” Persian complex predicates differ in lexicalization, and historical contexts may document a stage at which a construction remains analyzable in both ways.

A philologist should resist silently modernizing such a line to the first modern synonym that comes to mind. The historically available collocation, meter or cadence, and the manuscript's lexical choices are part of the evidence.

Manuscript spacing is evidence, not a parse

Premodern Persian manuscripts do not follow today's standardized conventions for word spacing, ZWNJ, می‌, plural suffixes, or compounds. A scribe may join material that a modern editor separates or separate material that modern orthography joins. Different witnesses can divide the same phrase differently.

Consequently, a diplomatic transcription should preserve what the witness shows, while a normalized edition should state its policy. Modern می‌کرد and خانه‌ها may be editorial normalizations even when the underlying morphology is secure. The same caution applies to the boundaries inside a complex predicate. The historical orthography page gives parallel principles for ezafe.

Avoiding false segmentation

Begin with the longest plausible lexical unit, then test it. روی آوردن به means “turn to/adopt”; parsing روی as an ordinary possessed “face” loses the construction. Conversely, not every noun before a frequent verb is a light-verb element: نامه را گرفت ordinarily means “received/took the letter,” with نامه as a referential object of heavy گرفتن.

Parallel contexts are crucial. If روی آوردن به repeatedly selects a goal with به and means adopting a practice, that distribution supports one lexical entry. If نامه varies freely with other concrete objects and accepts ordinary modifiers, heavy گرفتن is the better analysis.

Modern clitics preserve the old analytical problem

Discontinuity is not merely historical. In معرفی‌اش کردم, ـش is hosted by معرفی but is the object of معرفی کردن. In معرفی‌اش را خواندم, the same suffix is possessive inside the object noun phrase “his/her introduction.” The next verb and را distinguish the parses.

The dedicated clitic analysis supplies the full modern comparison. It also demonstrates why a word-by-word gloss can produce a false possessive reading.

Common Mistakes

1. Treating manuscript spacing as syntactic proof

2. Translating only the noun

3. Calling every noun plus frequent verb a compound

4. Modernizing without declaring it

Key Takeaways

• Syntactic or typographic separation does not prevent two elements from forming one lexical predicate.
• Diagnose unity through meaning, argument structure, lexical selection, inflectional scope, and parallel contexts.
• Classical order and manuscript spacing require witness-specific analysis; modern boundaries may be editorial.
• Test both errors: falsely splitting an idiom and falsely merging a referential object with a heavy verb.