An unfamiliar Lebanese verb should be diagnosed from a set of forms, not from one spelling. باع alone hides its short perfect stem and ī-class nonpast; بيوصل alone does not show perfect وصل wiṣel; حكى alone does not reveal حكيت. A one-form guess is especially unreliable because Arabic spelling records historical consonants and long vowels more clearly than everyday short vowels.
This page gives a repeatable workflow: collect principal parts, apply a suffix test, inspect related words, check an explicitly colloquial source, and reject any analysis that requires unattested forms. The goal is not merely to name a root class; it is to predict forms a Beirut speaker actually uses. The basic evidence card is introduced in Learning a Verb through Principal Parts.
Start with a five-part evidence card
For an unfamiliar verb, collect these forms before classifying it:
| Evidence | What it reveals | Example with باع |
|---|---|---|
| 3ms perfect | citation form and unsuffixed perfect stem | باع bēʿ |
| 1sg perfect | stem before a consonant-initial suffix | بعت biʿt |
| 3ms b-imperfect | nonpast stem and lexical vowel | بيبيع byibīʿ |
| 2ms imperative | command stem | بيع bīʿ |
| frame / related word | valency and root-family evidence | باع شي لحدا؛ بيع bayʿ |
بعت الطاولة لجاري، وهو رح يبيع كراسيه كمان.
biʿt eṭ-ṭāwle la-zhāre, w-huwwe raḥ ybīʿ krēsī kaman.
I sold the table to my neighbor, and he's going to sell his chairs too. (informal; short perfect beside bare ī-stem)
المحل بيبيع قطع قديمة، بس ما باع هالقطعة بعد.
el-maḥall byibīʿ ʾiṭaʿ ʾdīme, bass ma bēʿ hal-ʾiṭʿa baʿd.
The shop sells old pieces, but it hasn't sold this piece yet. (neutral; imperfect and long perfect evidence)
The suffix test: ask for “I did”
The first-person perfect is the fastest diagnostic because consonant-initial ـت forces hidden alternations into view.
| Citation → 1sg | Diagnosis | Evidence |
|---|---|---|
| كتب → كتبت katab → katabt | strong | basic stem stays stable |
| قال → قلت ʾāl → ʾelt | hollow | long middle vowel alternates with short stem |
| حكى → حكيت ḥaka → ḥkēt | final-weak | final edge rebuilds before suffix |
| طوى → طويت ṭawa → ṭawēt | doubly weak | medial glide remains while final edge rebuilds |
| حب → حبيت ḥabb → ḥabbēt | geminate, not weak | doubled radical expands before suffix |
قلت الحقيقة وحكيت كل التفاصيل.
ʾelt el-ḥaʾīʾa w-ḥkēt kill et-tafāṣīl.
I told the truth and related all the details. (neutral; hollow and final-weak suffix tests)
حبيت الاقتراح، بس طويت الورقة بالغلط.
ḥabbēt el-iʾtirāḥ, bass ṭawēt el-waraʾa bil-ghalaṭ.
I liked the proposal, but I folded the paper by mistake. (neutral; geminate versus doubly weak)
The test diagnoses a stem alternation, not an English tense. Every form in this section is perfect. Do not call قلت a “different past tense” from قال; person suffixes select different allomorphs of the same paradigm.
The nonpast test: find the lexical vowel
Remove only the outer marker after you have identified the actual b-imperfect. بيبيع → يبيع is a valid b/bare relation; an invented بيباع is not evidence. Compare the attested pairs:
- قال / بيقول ʾāl / byiʾūl
- باع / بيبيع bēʿ / byibīʿ
- نام / بينام nēm / byinēm
- وصل / بيوصل wiṣel / byūṣal
- حكى / بيحكي ḥaka / byiḥke
عادة بينام بكير، بس الليلة لازم ينام بعد الحفلة.
ʿāde byinēm bkīr, bass el-lēle lāzem ynēm baʿd el-ḥafle.
He usually sleeps early, but tonight he has to sleep after the party. (informal; same lexical stem inside b- and bare environments)
إذا ما في عجقة، بيوصل بكير؛ وإذا تأخر، رح يوصل عالسبعة.
iza ma fī ʿazhʾa, byūṣal bkīr; w-iza tʾakhkhar, raḥ yūṣal ʿas-sabʿa.
If there isn't traffic, he arrives early; if he's delayed, he'll arrive at seven. (informal; b-imperfect versus licensed bare form)
English tense labels do not perform this test. A translation such as “he goes” can correspond to a habit, schedule, vivid narrative, or other viewpoint. Diagnose the Lebanese stem first, then analyze b-, عم, رح, or the licensor.
Do not mistake a pattern vowel for a weak radical
A long vowel in the surface word does not automatically make a verb hollow. ساعد / بيساعد sāʿad / byisēʿed “help” is a derived Form III verb from s-ʿ-d; its long vowel belongs to the derived pattern. The suffix test ساعدت sāʿadet retains that pattern rather than producing a hollow short-stem alternation like قال → قلت.
ساعدت نادين تنقل الصناديق، وبعدين راحت عالبيت.
sāʿadet Nādīn tinʾol eṣ-ṣanādīʾ, w-baʿdēn rāḥet ʿal-bēt.
I helped Nadine move the boxes, and then she went home. (neutral; derived long vowel versus hollow motion verb)
Likewise, a visible و or ي may be a stable consonant, a weak radical, a prefix, or part of a suffix. Classification comes from correspondences across forms, not letter counting.
Related nouns recover roots—but not paradigms
Nouns and participles can reveal a radical obscured in the verb: قول supports q-w-l behind قال, بيع supports b-y-ʿ behind باع, and وصول supports initial w in وصل. This evidence is excellent for finding a lexical family.
وقت الوصول موجود عالتذكرة، والفعل هو وصل وبيوصل.
waʾt el-wuṣūl mawzhūd ʿat-tazkara, wil-fiʿel huwwe wiṣel w-byūṣal.
The arrival time is on the ticket, and the verb is wiṣel/byūṣal. (metalinguistic; noun confirms initial w)
But a related noun cannot safely generate the spoken verb. The noun بيع bayʿ does not tell you by itself whether the nonpast will contain ī, ū, or another vowel. Use relatives to propose a root, then verify the principal parts.
Dictionary lookup: verify variety, form, and frame
A useful lookup must explicitly document Levantine or Lebanese usage. Check that the entry gives vocalized or romanized pronunciation, at least a perfect/nonpast pair, and an example showing objects or prepositions. An MSA-only table can clarify spelling history but cannot establish an everyday Beirut form.
When sources disagree, ask what the disagreement represents:
- Is one form formal and the other colloquial?
- Is the variation Lebanese versus Syrian, Palestinian, or Jordanian?
- Is it phonetic variation in one paradigm, or a genuinely different stem?
- Does one source give a citation form while the other gives a suffixed form?
Record a regional form for recognition, but keep one coherent production baseline. Do not splice a Damascene perfect onto a Beirut b-imperfect and call the result “Levantine.”
A compact decision path
| Question | If yes | Next evidence |
|---|---|---|
| Does the 1sg shorten a medial long vowel? | suspect hollow | check 3ms b-imperfect vowel |
| Does final ى/ي rebuild before -t? | suspect final-weak | check command and clitic forms |
| Does initial w reshape after person prefixes? | suspect initial-weak | compare perfect and b-imperfect |
| Do two weak positions alternate? | suspect doubly weak | verify every principal part |
| Does a doubled radical expand before -t? | geminate, not weak | check perfect and imperative |
| Does hamza remain or change its seat? | hamzated | separate sound behavior from spelling |
Affirmative and negative minimal pairs
بيبيع هالمحل خرايط قديمة.
byibīʿ hal-maḥall kharāyeṭ ʾdīme.
This shop sells old maps. (affirmative habitual; verified ī-class)
ما بيبيع هالمحل خرايط قديمة.
ma byibīʿ hal-maḥall kharāyeṭ ʾdīme.
This shop doesn't sell old maps. (negative habitual; lexical stem and b- remain)
Dialogue: diagnosing from evidence
لقيت فعل طوى بالنص. كيف بعرف تصريفه؟
lʾēt fiʿel ṭawa bin-naṣṣ. kīf baʿref taṣrīfo?
Student: I found the verb ṭawa in the text. How do I know its paradigm? (metalinguistic dialogue)
دور على صيغة مع تاء وعلى المضارع.
dawwer ʿala ṣīgha maʿ tē w-ʿala l-moḍāreʿ.
Teacher: Look for a form with the t-suffix and for the nonpast. (instruction; traditional term used descriptively)
لقيت طويت وبيطوي، يعني في موقعين ضعاف.
lʾēt ṭawēt w-byiṭwe, yaʿne fī mawʾʿēn ḍʿāf.
Student: I found ṭawēt and byiṭwe, so two positions are weak. (evidence-based diagnosis)
مظبوط، بس بعد ناقصك الأمر وطريقة أخد المفعول.
maẓbūṭ, bass baʿd nāʾṣak el-amer w-ṭarīʾet akhd el-mafʿūl.
Teacher: Correct, but you still need the imperative and the way it takes an object. (metalinguistic; paradigm plus frame)
Common Mistakes
Guessing from one perfect form
❌ بما إنو باع فيه ألف، المضارع أكيد بيباع.
Incorrect — written alif does not determine the nonpast class.
✅ باع وبعت وبيبيع مجموعة لازم تنحفظ سوا.
bēʿ w-biʿt w-byibīʿ mazhmūʿa lāzem tinḥafaẓ sawa.
Bēʿ, biʿt, and byibīʿ are a set that must be memorized together. (metalinguistic)
Treating every long vowel as hollow
❌ ساعد فعل أجوف لأنه فيه ألف.
Incorrect — ساعد is a derived Form III verb; the long vowel is part of that pattern.
✅ ساعد وساعدت وبيساعد بيكشفوا جذر قوي.
sāʿad w-sāʿadet w-byisēʿed byikshfo zhizer ʾawe.
Sāʿad, sāʿadet, and byisēʿed reveal a strong root. (metalinguistic)
Using an MSA dictionary as proof of Beirut speech
❌ لقيت الشكل بالفصحى، يعني هو نفسه بالحكي اللبناني.
Incorrect — formal evidence can identify a root without verifying the Lebanese principal parts.
✅ بتأكد من مصدر لبناني وبسجل اللفظ والسياق.
btʾakkad min maṣdar libnēne w-basazhzhil el-lafẓ wis-siyēʾ.
I verify it in a Lebanese source and record the pronunciation and context. (academic learning strategy)
Regularizing after identifying the class
❌ عرفت إنه ضعيف، ففيني اخترع باقي الصيغ.
Incorrect — a class narrows the possibilities but does not supply lexical vowels or suppletive forms.
✅ التصنيف بيساعدني اسأل الأسئلة الصح، مش اخترع الجواب.
et-taṣnīf byisēʿedne isʾal el-asʾile ṣ-ṣaḥḥ, mish khtareʿ el-zhawēb.
Classification helps me ask the right questions; it doesn't let me invent the answer. (academic explanation)
Forgetting the complement frame
❌ حفظت التصريف وما عرفت بأي حرف جر بستعمل الفعل.
Incomplete learning — morphology alone does not show how participants attach.
✅ بسجل باع شي لحدا مع التصريف.
basazhzhil bēʿ shi la-ḥada maʿ et-taṣrīf.
I record ‘sell something to someone’ together with the paradigm. (metalinguistic)
Key Takeaways
Diagnose an unfamiliar verb from a network: 3ms and 1sg perfect, b-imperfect, imperative, related words, and complement frame. Use the suffix test to expose hidden stems, but distinguish weak alternations from derived long vowels and geminate expansion. Root-family evidence proposes an analysis; a variety-specific dictionary or corpus verifies it. Classification should guide better questions, never authorize invented Beirut forms.
Now practice Arabic
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Arabic →Related Topics
- Learning a Verb through Principal PartsA2 — Learn the small set of forms that reveals a Lebanese verb's stems, irregularities, commands, and complement pattern.
- Weak Verbs: An OverviewB1 — Recognize initial, hollow, final, and doubly weak Lebanese verbs and learn them reliably through principal parts rather than spelling guesses.
- Initial-Weak VerbsB1 — Learn how an initial w or y behaves across Lebanese perfect, imperfect, command, participle, and maṣdar forms without importing an MSA deletion rule.
- Hollow VerbsB1 — Understand why Lebanese hollow verbs alternate between long-vowel and short-vowel stems across the perfect, imperfect, and imperative.
- Final-Weak VerbsB1 — Recognize how Lebanese vowel-final verbs rebuild their stems before person endings, clitics, and imperfect markers.
- How to Find a Root in a Spoken WordB2 — Learn a cautious, evidence-based method for recovering consonantal roots from real Lebanese words.