How Somali packages meaning
Somali grammar at a glance
At a glance
| Information type | Plain question | How Somali encodes it |
|---|---|---|
| Actor | Who does the action? | SUFFIXSEPARATE WORD |
| Recipient | Who or what receives the action? | WORD ORDERSEPARATE WORD |
| Time | When does the action happen? | SUFFIXSEPARATE WORD |
| Aspect | Finished or still happening? | SUFFIXSEPARATE WORD |
| Certainty / mood | Real fact, or desired / possible / hypothetical? | SUFFIXSEPARATE WORD |
| Negation | Did it not happen? | SEPARATE WORDSUFFIX |
| Voice | Is the subject doing, being done to, or causing? | SUFFIXSEPARATE WORD |
| Evidentiality | How does the speaker know this? | CONTEXT |
| Specificity | This particular one, or any one? | SUFFIX |
| Quantity | How many? | SUFFIXREDUPLICATIONALTERNATIONCONTEXT |
| Possession | Whose is it? | SUFFIXTONES |
| Location / Space | Where? In which direction? | SEPARATE WORDWORD ORDER |
| Information structure | What is this sentence about? What is new? | SEPARATE WORDWORD ORDERTONES |
| Register | What social relationship does the speaker encode? | CONTEXTWORD ORDER |
How it stacks
A single article suffix on the noun fuses three things at once: gender (via /k/ vs. /t/), number (via article shape after the plural stem), and a time/discourse reference — "near" -ka/-ta for now or generic, "remote" -kii/-tii for past or previously-mentioned. Changing any of the three requires picking a different article.
The same /k/~/t/ pair reappears all over the grammar — in demonstratives (-kan/-tan "this"), possessives (-kayga/-tayda "my"), and interrogatives (kee/tee "which one?"). The underlying /k/ and /t/ undergo regular alternations (e.g. after /l/, /t/ → /sh/, so walaal+ta → walaasha "the sister").
In detail
Actor in Somali — Who does the action?
Person and number are encoded both by a suffix on the verb AND by a short subject pronoun (aan, aad, uu, ay...) that sits inside the "particle phrase" before the verb. Full independent pronouns (aniga "I", adiga "you"...) are optional — used only for emphasis. Somali distinguishes inclusive "we" (aynu — including you) from exclusive "we" (aannu — not including you).
Verb suffixes: 1SG and 3SG.M share one form (-aa), while 2SG and 3SG.F share another (-taa) — a systematic syncretism that Somali inherits from its Cushitic ancestry and shares with Oromo.
Short subject pronouns contract obligatorily with the focus particle: waa + aan → waan (1SG), waa + uu → wuu (3SG.M), waa + ay → way (3SG.F/3PL), baa + aan → baan, etc.
Subjects drop their lexical High tone when they are non-focused — this is how the grammar "marks" them beyond any suffix (see Recipient/case details).
Five common verbs preserve the archaic Afroasiatic prefix conjugation: yahay "is", yaallaa "is located", yaqaannaa "knows", yimaaddaa "comes", yiraahdaa "says". The copula yahay inflects as ahay "I am", tahay "you are / she is", yahay "he is", nahay "we are", with the person markers /n-, t-, y-/ appearing as PREFIXES, not suffixes.
Compare: Spanish hablo / hablas already tells you who does it. Somali also has verb endings, but layers on TWO more signals — a short subject clitic in the particle phrase AND a tone shift on the subject noun — making "who acts" triply encoded.
Recipient in Somali — Who or what receives the action?
Direct objects are unmarked (absolutive — the bare dictionary form). The INDIRECT object / recipient is signalled by one of four "preverbal prepositions" — u, ku, ka, la — that sit just before the verb, not before the noun. Short object pronouns (i, ku, na, ina, idin, is) also sit in the verb phrase; there is famously NO overt 3rd-person object pronoun.
The four preverbal prepositions are: u "to/for", ku "in/at/with (instrument)", ka "from/about", la "with (company)". They cluster in a fixed order inside the verb phrase and obligatorily fuse together: u+ku → ugu, ku+ka → kaga, ku+la → kula. The grammar does NOT tell you which NP each preposition applies to — context disambiguates.
Object pronouns (i, ku, na, ina, idin, is) also sit in the verb phrase, in a fixed slot before the prepositions. "is" is the reflexive (covers all persons).
There is no 3rd-person object pronoun at all — to say "he hits him", one simply says "wuu garaacaa" ("he hits"), and context picks out who.
Ditransitive verbs like sii "give", tus "show", weydiiyi "ask" can take two direct objects without any preposition — only the recipient argument of other verbs is signalled with u.
Compare: English "give the book to Hassan" puts "to" right before Hassan. Somali keeps "u" stuck to the verb — you say, roughly, "Hassan book to-gives" with "to" sitting next to the verb, not next to "Hassan".
Time in Somali — When does the action happen?
Present and past are marked by a vocalic alternation in the verb suffix: present -aa / past -ay (also -taa/-tay for 2SG+3SG.F). Future is periphrastic: the infinitive plus the auxiliary verb doon- "want" in the present tense. Near-future events routinely use the present progressive instead.
The future construction is historically a grammaticalised "want to X" — a very widespread source of future tenses (compare English "will", which is cognate with German wollen "want").
The vowel alternation is regular: present -aa / past -ay, present -taa / past -tay, present -naa / past -nay, present -taan / past -teen, present -aan / past -een.
Time adverbs (shalay "yesterday", berri "tomorrow", maanta "today") are frequent and often appear even when the verb is already tense-marked.
Compare: Mandarin marks tense with separate time words and leaves the verb unchanged. Somali packs the tense into the verb suffix — changing just one vowel flips present to past.
Aspect in Somali — Finished or still happening?
Progressive aspect inserts -ay- between the stem and the tense suffix (heesayaa "is singing" ← hees "sing" + ay + aa "PRS"). Habitual past uses the infinitive plus an inflected form of jir- "exist/be present" (heesi jiray "used to sing"). The simple past carries perfective meaning by default.
The progressive infix -ay- descends historically from a fused auxiliary construction hees + haya "holds singing" — similar to how English "is -ing" evolved from "is on singing".
The habitual past auxiliary jiray is the past of the verb jir- "to exist, be present" — "used-to-sing" is literally "singing existed".
Somali does NOT separately grammaticalise the perfective/imperfective contrast the way Russian does — the simple tenses cover both.
Compare: English forms progressive with the separate words "be" + "-ing". Somali fuses both into the verb — a single suffix -ay- captures "in progress".
Certainty / mood in Somali — Real fact, or desired / possible / hypothetical?
Mood is encoded by a combination of the sentence-particle slot (má negator, ma question, ha prohibitive, há optative) and the verb suffix (indicative -aa/-ay vs. subjunctive -o vs. reduced -in). The subjunctive -o appears after negation, in positive subordinate clauses with a distinct subject, and in wishes and indirect commands. Conditional clauses are headed by haddii "if" plus either subjunctive or the conditional construction with lahaa "would".
The optative particle há (with High tone) and the prohibitive ha (toneless) are minimal pairs distinguished only by pitch — a reminder that tone in Somali is fully grammatical.
Subjunctive -o replaces indicative -aa after negation (Má heeso "She does not sing"), in positive subordinate clauses with an expressed subject, and in wishes. It is NOT a "speaker doubt" marker — it is driven by clause type.
The conditional construction uses lahaa (from leh "having" + yahay "is") with an infinitive — literally "had singing". It covers both English "would sing" (hypothetical) and "would have sung" (counterfactual); context disambiguates.
There is no dedicated "modal verb" class — meanings like "must", "should", "might" are expressed lexically (waa in... "it is that..." = must) or through the conditional.
Compare: English uses separate modal verbs (will, would, must, can) before a bare verb. Somali uses a combination of a sentence-type particle (má, ha, há) and a verb-form suffix (-aa, -o, -in) — the two together pick out the mood.
Negation in Somali — Did it not happen?
Negation uses the sentence-particle má "not" (with High tone) in place of a focus particle, and it forces the verb into the subjunctive. A different negator aan appears in subordinate clauses and in main clauses that already carry a focus particle — aan takes a reduced verb ending in -in / -n. Negation thus REPLACES the focus particle rather than stacking with it.
má carries a High tone and replaces the focus particle position — one or the other, never both in the same clause.
aan is the "subordinate / post-focus" negator and it takes the reduced subjunctive ending -in / -n (heesin, cunin, ordayn). It is used in relative and complement clauses, and in main clauses that already have baa/ayaa/waxaa.
There is no double negation — a single particle plus a verb-form shift does the job.
Compare: French puts "ne...pas" around the verb and piles on more negators (ne...jamais, ne...personne). Somali has a single negator (má or aan, picked by clause type), and negation replaces the positive clause-type marker entirely.
Voice in Somali — Is the subject doing, being done to, or causing?
Voice is built by stacking derivational suffixes on the verb root: causative -i, -eeyaa, -siiyaa (with increasing indirectness), middle/anticausative -maa, autobenefactive -(s)taa, and reflexive via the object clitic "is". There is no dedicated passive morphology — the impersonal subject clitic "la" ("one, they") serves the same function: la akhriyaa "it is read" (lit. "one reads").
The causative has three levels of indirectness: -i (direct: "feed" ← "eat"), -eeyaa (factitive: "make white" ← "white"), and -siiyaa (indirect, from historical sii "give"). Stacking is productive.
The "passive" function is handled by the impersonal subject clitic la "one/they" — the verb stays in the active form; Somali simply demotes the agent by using an unspecified "one".
The reflexive is "is" (covers all persons) and sits in the object-clitic slot inside the verb phrase.
Compare: English "was read" uses a separate auxiliary plus a past participle. Somali does not bother with a true passive — it says "one reads the book", with the impersonal "la" replacing the agent entirely.
Evidentiality in Somali — How does the speaker know this?
Somali does not grammaticalise evidentiality. There are no obligatory verb markers for witnessed vs. hearsay vs. inferential information. Speakers use lexical strategies — adverbial phrases and quotative verbs — to signal the source of information.
Even though Somali has a rich tense/aspect system, it does not slot in a witnessed-vs-hearsay contrast the way Turkish (-di vs. -miş) or Bulgarian does.
Typical hearsay expressions use the impersonal: la yidhi "(someone) said"; waxa la sheegay "what one reported"; sida la yidhi "as is said".
Compare: Turkish obligatorily distinguishes "I saw him leave" (yaptı) from "I heard he left" (yapmış) in the verb ending. Somali does not — you would add an adverbial phrase only if you specifically wanted to flag hearsay.
Specificity in Somali — This particular one, or any one?
Definiteness is marked by a suffix on the noun that fuses gender with a time/discourse distinction: near/generic -ka (masc) / -ta (fem) for "the", and remote -kii / -tii for "the one we know about" or "that past one". There is NO indefinite article — a bare noun is indefinite. Demonstratives (-kan/-tan "this", -kaas/-taas "that") are the same paradigm with different vowels.
The near/remote article distinction is a real two-way discourse system — NOT the same as "definite vs. indefinite". Both -ka and -kii are definite; they differ in whether the referent is present/generic (near) or previously-mentioned/past (remote).
Article shape varies predictably by the preceding sound: after /g d/ geminate (-gga, -dda); after /e o/ a filler /h/ (aabbe "father" → aabbaha "the father"); after /l/ the /t/ fuses to /sh/; after pharyngeals the /k/ drops (magac "name" → magaca "the name").
A bare plural is also indefinite: buugag "books" vs. buugagga "the books".
To say "a certain / some book", Somali uses the numeral/indefinite determiner hal ("one") or a post-nominal qualifier.
Compare: English picks between "a/an" and "the" but has only one "the". Somali skips the indefinite altogether and splits "the" into a TWO-way near/remote contrast that carries a time reference — "last week" vs. "the week (in general)" are distinguished by the article alone.
Quantity in Somali — How many?
Plural formation is notoriously diverse — Saeed distinguishes seven declension classes by plural shape (suffix -o, reduplicating -aC, -oyin, -yaal, Arabic broken plurals, tone-only). Most classes show gender polarity: a masculine singular pluralises as feminine, and vice versa. After numerals greater than one, the noun appears in a "counting form" — for one class (feminine -o plurals) this is a distinct plural-genitive -ood; for every other class the counting form is identical to the singular.
The seven declensions cross three axes: polarity direction (F→M, M→F, or none), plural shape (-o/-yo/-oyin/-yaal/reduplication/Arabic/tone-only), and tone pattern (H on last vs. penultimate mora). Learners should memorise plural + article per word.
Gender polarity is the rule, not an exception: bisad (F) "cat" → bisado (M) "cats"; buug (M) "book" → buugaag (F) "books"; gabadh (F) "girl" → gabdho (M) "girls". But the reduplicating class (D4: nin/niman "man/men", af/afaf "mouth/mouths") does NOT show polarity — these stay masculine.
The counting form is not a separate plural — it is a noun in a singular or plural genitive after a numeral head. Only the feminine -o class has a distinct plural-genitive -ood; all other classes fall back to the singular.
Inclusive "we" (aynu) vs. exclusive "we" (aannu) — a subject-pronoun distinction also relevant to quantity of participants.
Compare: English simply adds -s ("cat/cats") with a stable gender (or no gender) and a stable article. In Somali, pluralising a noun often changes its gender and thus its article — bisad-da "the cat" (fem) becomes bisad-aha "the cats" (masc, via F→M polarity).
Possession in Somali — Whose is it?
Pronoun possession fuses into the definite article: -kayga/-tayda "my", -kaaga/-taada "your", -kiisa/-tiisa "his", -keeda/-teeda "her", and so on — gender-agreeing suffixes on the noun. Noun possession uses a head-first genitive: possessed noun + possessor, where the possessor carries a tonal shift (H moves to the last mora) and optionally a segmental suffix (-eed, -aad, -ood) depending on its declension.
Possessive suffix paradigm: 1SG -kayga/-tayda; 2SG -kaaga/-taada; 3SG.M -kiisa/-tiisa; 3SG.F -keeda/-teeda; 1PL.EXCL -kayaga/-tayada; 1PL.INCL -keenna/-teenna; 2PL -kiinna/-tiinna; 3PL -kooda/-tooda. Each fuses the possessor with the head noun's gender.
Close-kin "short forms" drop the terminal article vowel: hooyaday "my mother", aabbahay "my father", walaalkay "my brother". With a modifier the full form returns.
Noun genitive is head-first — possessed first, then possessor. A tone shift marks the possessor even when there is no segmental suffix; the -eed/-aad/-ood suffixes appear only in specific declension classes.
For standalone "mine / yours", a separate pronoun-possessive combines with kan / tan: kan kayga "mine (masc.)", tan tayda "mine (fem.)".
Compare: English "my book" uses a separate word "my". Somali suffixes the whole possessor onto the noun — buug-kayga — and the suffix even encodes the gender of what is owned, not just of the owner.
Location / Space in Somali — Where? In which direction?
Somali has NO prepositions before nouns and NO locative case suffix on nouns. Instead it has four "preverbal prepositions" — u "to/for", ku "in/at/with", ka "from/about", la "with (company)" — that cluster just before the verb. Fine-grained space comes from postpositions/position nouns (hoos "under", dul "above", hor "in front", gees "beside") placed AFTER the noun, often in combination with a preverbal preposition. Two deictic particles (soo "hither/ventive", sii "thither/itive") slot inside the verb phrase to signal direction of motion relative to the speaker.
The four prepositions u, ku, ka, la sit in a fixed slot inside the verb phrase (roughly: between object clitics and the verb). They fuse obligatorily when stacked: u+ku → ugu, ku+ka → kaga, ka+la → kala, la+u → loo. The cluster takes exactly one High tone on its last element.
Position nouns (dul, hoos, hor, dhex, kor, ag) follow the noun they relate to — they are postpositions in form. Often they co-occur with a preverbal preposition: miiska dul saaran "placed on top of the table" = table above put-on.
Somali has no morphological locative case — all "location" relationships are expressed through the preverbal-preposition system, position nouns, and deictic particles.
The deictic particles soo (ventive — toward speaker or reference point) and sii (itive — away from speaker) are a striking Cushitic feature; they slot in the verb phrase and can combine with u / ka for fine-grained direction.
Compare: English "in the house" puts "in" RIGHT before "the house". Somali puts "ku" (in/at) not before "guriga" but at the end of the clause, just before the verb — so the sentence sounds to a Western ear like "the house he at-is".
Information structure in Somali — What is this sentence about? What is new?
Somali is defined by an OBLIGATORY focus system: every positive declarative main clause must contain exactly one focus particle, and which one it is changes the meaning. waa focuses the verb/predicate (general declarative); baa or ayaa focuses the immediately preceding NP (narrow NP focus, "it is X who..."); waxa(a) focuses an NP that comes AFTER the verb (sentence-final cleft). A focused subject loses its nominative marking, loses its subject clitic, and forces the verb into a reduced paradigm.
Every positive main declarative MUST carry exactly one of waa / baa / ayaa / waxa(a). A clause without a focus particle is either a command, a question with ma, a subordinate clause, or negated (all of which replace the focus slot with their own particle).
Subordinate clauses have no focus particle — they drop back to neutral order.
When the subject is focused (with baa or ayaa), three things happen at once: the subject drops its nominative case marking (no -u/-i suffix), the subject clitic disappears from the particle phrase, and the verb switches to the "reduced" paradigm (three-way distinction: one form for 1SG/2SG/3SG.M/2PL/3PL, one for 3SG.F, one for 1PL).
Wh-words obligatorily fuse with baa/ayaa: ayo + baa → yaa "who (focused)"; maxay + baa → maxaa "what (focused)".
The difference between waa and the NP-focus particles is whether the NEW information is the whole verb/event (waa) or a specific noun phrase (baa/ayaa/waxaa).
Compare: English marks focus mostly with prosodic stress ("ALI is reading") or an optional cleft ("It is Ali who is reading"). In Somali the focus machinery is OBLIGATORY and lexical — you literally cannot say the sentence without picking a focus particle first.
Register in Somali — What social relationship does the speaker encode?
Somali does not have a full honorific or T/V system built into the grammar. Politeness is expressed pragmatically — through address terms (aabbe "father", walaal "sibling", macallin "teacher"), poetic and Quranic phrasing, and elaborate greeting rituals. The only systematic register contrast is the 2PL form being usable as a "formal singular" in some contexts and the literary / poetic register (gabay) drawing on archaic vocabulary and alliterative morphology.
Unlike Spanish (tú/usted), German (du/Sie), or Japanese (plain/polite/humble), Somali has no grammatical pronoun split for respect. The 2PL idinka can soften address to one person, but it is not a dedicated formal pronoun.
Politeness is cultural and lexical: titles (macallin "teacher", sayid "master"), kinship terms used to non-relatives (walaal "brother/sister", eedo "auntie"), and formulaic blessings (Alla ha u naxariisto "may God have mercy on him").
The literary / oral-poetry register (gabay, buraanbur) uses archaic vocabulary and alliterative morphology — but this is a stylistic register, not a grammaticalised politeness tier.
The vocative shifts tone to the first mora (Máxamed!) and can add dedicated suffixes: -ow (masc.), -ey/-aay/-ooy (fem.), -yahow / -yahay (plural).
Compare: Japanese encodes register in nearly every sentence — plain, polite -masu, humble, honorific verb forms. Somali leaves register almost entirely outside the grammar, signalling respect through word choice, titles, and formulas instead.
Common questions about Somali
What is marked-nominative alignment?
What's the difference between inclusive and exclusive 'we' in Somali?
Is Somali tonal?
Does Somali have grammatical gender?
Why does Somali have two ways to say 'the'?
Sources for Somali
The grammatical descriptions on this page are informed by the following published reference and descriptive grammars. Grammatical facts themselves are not subject to copyright; the scholars who documented them deserve attribution.
- Saeed, John Ibrahim (1999). Somali. Descriptive Grammars series. London: Routledge. [THE primary reference grammar for Somali]
- Saeed, John Ibrahim (1993). "Focus, topic and discourse coherence in Somali." Journal of African Languages and Linguistics 14(2): 191–212.
- Andrzejewski, B. W. and I. M. Lewis (1964). Somali Poetry: An Introduction. Oxford: Clarendon Press.
- Orwin, Martin (1995). Colloquial Somali. London: Routledge.
- Zorc, R. David and Mohamed Osman (1993). Somali-English Dictionary. Kensington, MD: Dunwoody Press.