Persian has three visible boundary choices inside expressions that English may call “one word”: no boundary (دانشگاه), a zero-width non-joiner or ZWNJ (دانشآموز), and an ordinary space (تصمیم گرفتن). These are not three interchangeable typographic styles. The choice reflects lexical convention, morphology, readability, and—in compound verbs—the fact that one lexical predicate can still contain two orthographic words.
The Academy's spelling principles establish obligatory cases but deliberately leave some compounds to accepted usage. Consequently, there is no single mechanical rule that predicts every lexicalized compound. A careful writer combines productive rules with dictionary evidence, then treats real variants as a search problem rather than pretending they do not exist. Review the mechanics first on Spaces, half-spaces, and the ZWNJ.
Three boundary types
| Boundary | What is stored | Typical function | Examples |
|---|---|---|---|
| solid | adjacent base letters | strongly lexicalized words and tightly bound formatives | دانشگاه، کتابخانه، روزنامه |
| ZWNJ | U+200C between parts | one orthographic word whose visible parts should not join | دانشآموز، تصمیمگیری، بهکارگیری |
| space | U+0020 between words | multiword expressions and the parts of compound verbs | هیئت علمی، تصمیم گرفتن، کار کردن |
A ZWNJ is still present after a non-connecting letter even when it causes no obvious visual change. کارها and کارها may look alike because ر already breaks joining, but only the second string explicitly contains the boundary. Visual sameness therefore does not imply data identity.
Solid spelling: lexical unity wins
Many established compounds have become conventional single strings. دانشگاه “university” and کتابخانه “library” are internally analyzable, but contemporary readers retrieve each as a lexical unit. Productive suffix-like elements such as ـگاه and ـخانه frequently form solid words when convention supports it.
کتابخانهی دانشگاه تا ساعت هشت باز است.
ketâbkhâne-ye dâneshgâh tâ sâ'at-e hasht bâz ast.
The university library is open until eight.
این روزنامه نسخهی دیجیتال هم دارد.
in ruznâme noskhe-ye dijitâl ham dârad.
This newspaper also has a digital edition.
Solid spelling does not mean that every possible combination with گاه or خانه may be fused by analogy. Lexicalization is partly conventional, and dictionaries record that convention. This is one of the genuinely memorized areas of Persian orthography.
ZWNJ: one word, protected internal boundary
The ZWNJ blocks cursive joining without creating the full separation of a space. It is particularly useful where both sides remain recognizable: prefixes such as می/نمی, productive endings such as ها, and many compounds and derivatives.
دانشآموزها نتیجهی پژوهش را ارائه کردند.
dânesh-âmuz-hâ natije-ye pazhuhesh râ erâ'e kardand.
The students presented the research findings.
تصمیمگیری بدون اطلاعات کافی دشوار است.
tasmim-giri bedun-e ettelâ'ât-e kâfi doshvâr ast.
Decision-making without enough information is difficult.
بهکارگیری این روش زمان را کم میکند.
be-kârgiri-ye in ravesh zamân râ kam mi-konad.
Using this method reduces the time required.
Notice the hierarchy in دانشآموزها: one ZWNJ divides the compound internally, while another records the plural boundary after ز even though that non-connecting letter already creates a visible break. In تصمیمگیری, the boundary protects both lexical parts. The romanization shows such productive boundaries with a hyphen; the Persian line uses no visible hyphen.
A full space: more than one orthographic word
An ordinary space separates syntactic words. It is therefore required in phrases such as هیئت علمی “academic staff” and in Persian compound verbs, where a noun, adjective, or prepositional element combines lexically with an independently inflected light verb.
هیئت علمی پیشنهاد تازه را بررسی کرد.
hey'at-e elmi pishnehâd-e tâze râ barrasi kard.
The academic staff reviewed the new proposal.
پس از جلسه تصمیم گرفتیم صبر کنیم.
pas az jalase tasmim gereftim sabr konim.
After the meeting, we decided to wait.
خواهرم در یک شرکت کوچک کار میکند.
khâharam dar yek sherkat-e kuchek kâr mi-konad.
My sister works at a small company.
Writing تصمیمگرفتیم would falsely turn the boundary between the nonverbal element and the finite verb into an internal word boundary. The predicate is lexical as a whole, but گرفتیم still carries tense and person as its own orthographic word. See Compound verbs as single lexical predicates for the grammatical consequences.
How Academy norms actually work
Academy standardization aims at consistency without claiming that every compound can be derived from a single formula. Strong lexicalization favors solid spelling. A productive, still-visible boundary often favors ZWNJ. Genuine syntactic separation requires a space. Additional pressures include avoiding an unreadable mass of teeth-like strokes, preventing ambiguity, and preserving a word's established public spelling.
That leaves a middle zone. Dictionaries, publishers, and older editions may disagree over a compound, and current house styles may select one admitted form. For professional work, consult the latest Academy spelling dictionary or the publication's style sheet. Do not “correct” a proper name against its legally or historically established form.
Search across variants without corrupting the source
Search engines and corpora differ in whether they equate a space, ZWNJ, and no boundary. When retrieval matters, query the canonical spelling first, then controlled variants: دانشآموز, دانش آموز, and دانشآموز. Also test Arabic/Persian kaf and yeh variants in legacy data. This broad search is for discovery; your edited text should still store the chosen standard form.
همهی صورتهای املایی را در نمایه ثبت کردیم.
hame-ye surat-hâ-ye emlâyi râ dar namâye sabt kardim.
We recorded all the spelling variants in the index.
In a scholarly edition, keep the source string and a normalized search key in separate fields. Otherwise, improving searchability can silently falsify the evidence you meant to preserve.
Common Mistakes
1. Replacing every ZWNJ with a space
❌ دانش آموزها رسیدند.
Incorrect in the selected standard — a full space splits دانشآموز, and the plural boundary is also missing.
✅ دانشآموزها رسیدند.
dânesh-âmuz-hâ residand.
The students arrived.
2. Joining the two parts of a compound verb
❌ تصمیمگرفتم فردا بروم.
Incorrect — the nonverbal element and finite light verb are separate orthographic words.
✅ تصمیم گرفتم فردا بروم.
tasmim gereftam fardâ beravam.
I decided to go tomorrow.
3. Fusing a familiar solid word inconsistently
❌ کتاب خانه بسته است.
Incorrect here — the lexical item ketâbkhâne ‘library’ is conventionally solid.
✅ کتابخانه بسته است.
ketâbkhâne baste ast.
The library is closed.
4. Treating an invisible boundary as optional data
❌ بچهها در حیاط بازی میکنند.
Incorrect — the plural suffix needs a ZWNJ after بچه.
✅ بچهها در حیاط بازی میکنند.
bache-hâ dar hayât bâzi mi-konand.
The children are playing in the courtyard.
5. Adding a long-vowel letter for an unwritten short vowel
❌ کیتابخانه امروز باز است.
Incorrect — the short e in ketâb is not written with ی.
✅ کتابخانه امروز باز است.
ketâbkhâne emruz bâz ast.
The library is open today.
Key Takeaways
Use solid spelling for established lexical units, ZWNJ for protected boundaries inside one orthographic word, and a space between syntactic words—including the parts of compound verbs. Productive rules cover much of the system, but lexical convention remains real. Search tolerant variants when necessary, then store one deliberate standard form and preserve any historical source form separately.
Now practice Farsi
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Farsi →Related Topics
- Spaces, Half-Spaces, and the Zero-Width Non-JoinerA2 — Learn how Persian distinguishes word spaces, invisible half-spaces, and natural cursive breaks inside a word.
- The Prefixes می and نمیB1 — Type the imperfective prefixes می and نمی with ZWNJ, place them correctly in simple and compound verbs, and recognize colloquial spellings.
- Plural ها and Comparative تر with ZWNJB1 — Write productive ها, تر, and ترین boundaries correctly, including ezâfe forms, vowel-final hosts, and lexicalized exceptions.
- Orthographic Variation before and after Academy StandardizationC1 — Recognize older Persian spellings, apply modern Academy conventions, and distinguish faithful transcription from normalization.
- Noun-Noun and Noun-Stem CompoundsB1 — Analyze Persian noun compounds through head direction, verbal-stem relations, stress, lexical spelling, semantic opacity, and outer-edge plural marking.
- Compound Verbs as Single Lexical PredicatesA1 — Learn why Persian noun- or adjective-plus-verb combinations function as one predicate and must be stored with their light verb and complement pattern.