Tchernichovsky: אני מאמין (שחקי, שחקי)

The poem's title is אֲנִי מַאֲמִין. Everyone calls it שחקי, שחקי after its opening words, and that nickname is now more famous than the title, but the title is what Tchernichovsky gave it. He wrote it in Odessa in 1892 and printed it in his first collection, חזיונות ומנגינות (1898); he died on 14 October 1943, so the poem is public domain and quotable entire. This page quotes two of its eight stanzas — the first and the sixth — because those two happen to contain almost every grammatical oddity worth a C2 reader's attention, and printing the other six would add scenery rather than grammar. The Naomi Shemer page in this group, by contrast, is limited to two lines by copyright.

How the text is set, and what was checked

The pointed reading follows the text at Zemereshet, whose stated source is the ברון יחד songbook (1963) collated against the 1937 כל שירי שאול טשרניחובסקי held by the National Library; he.wikipedia's pointed inset of the poem agrees with it word for word, and its article on Tchernichovsky quotes the sixth stanza independently in unpointed form. The displayed line is that pointed text's consonantal skeleton. In these two stanzas that skeleton happens to coincide with modern ktiv male at every single word — unlike the Bialik and Rachel pages in this group, nothing here changes spelling between the classical text and a modern one. The one word you would never see in a modern text is מנו, and that is a matter of register rather than orthography. Editions differ in whether they set the compounds with a maqaf or a hyphen; this page uses the maqaf.

The poem's stress is mil'el throughout, on the Ashkenazi pattern of its period, which is how the melody scans. The romanizations below give the modern Israeli reading, which is milra; that mismatch is real and worth knowing about before you sing along.

The opening stanza

שַׂחֲקִי is two verbs at once

It is written with שׂ (sin), not shin, and the root is שׂ־ח־ק — the same root as Ecclesiastes' עֵת לִשְׂחוֹק, "a time to laugh", on the Ecclesiastes 3 page. Here it is a pa'al feminine singular imperative of an ayin-guttural root: the ח cannot take a plain shva, so it takes a chataf-patach, and the vowel in front of it is drawn to patach in sympathy.

And now the part nobody tells you. Run the two binyanim of this root side by side:

The masculine forms are different words. The feminine forms are the same word, letter for letter and vowel for vowel (the pi'el paradigm is on Pealim as שַׂחֵק / שַׂחֲקִי; the pa'al pattern is the same one Pealim gives for the shin-verb שָׁחַק "erode", שְׁחַק / שַׁחֲקִי). So the line that opens the most famous Hebrew poem of national hope is formally identical to what an Israeli parent says to a child at a playground:

And for plain laughing, modern Hebrew has moved to a third root entirely:

זוּ is not "this"

זוּ in line 2 looks like the feminine demonstrative זוֹ and is not. It is an archaic relative particle, equivalent to אֲשֶׁר or modern ש־ — "the dreams that I, the dreamer, tell". Biblical Hebrew uses it that way in the Song of the Sea:

This is the stanza's real trap, and the reason it is worth naming: read זוּ as "this" and line 2 becomes an unparseable fragment. He.wikipedia's own annotation of the poem flags the word for exactly this reason.

אַאֲמִין, with two alefs

אַאֲמִין is the first-person future of the hif'il הֶאֱמִין "to believe", root א־מ־ן. Because the future prefix for "I" is also an א, the word begins with two of them in a row — the prefix alef carrying patach, the root alef carrying a chataf-patach. Nothing has been doubled by mistake. Modern Hebrew writes and says exactly this form; it is only the poem's literary choice of a bare future where speech would use the present participle מאמין that dates the line.

עוֹדֶנִּי, and a live corner of the grammar

עוֹדֶנִּי is עוֹד carrying a first-person suffix: "I still…". Particles that inflect like this are the thing textbooks demote to a footnote, and they are not marginal at all — they are how formal Hebrew negates, presents and emphasises:

באדם אאמין is a statement

Line 3 puts the prepositional complement in front of the verb. English uses fronting so often for questions ("in whom do you believe?") that learners read this as interrogative. It is not — Hebrew marks a yes/no question with האם or with intonation, never with word order, so fronting is free to do what it does everywhere else in the language: contrast and emphasis.

The sixth stanza

יָשׁוּב יִפְרַח: two finite verbs, one event

יָשׁוּב יִפְרַח is not "it will return and it will bloom". It is a serial construction in which שׁוּב has been bleached into an auxiliary meaning "again": it will bloom again. Two finite verbs, same subject, same tense, no conjunction between them.

Modern Hebrew kept the construction and added a vav: שב ו־ and חזר ו־ followed by a second finite verb mean "did it again", and neither is a compound verb you will find in a beginner's textbook. On he.wikipedia (insource:/ חזר והדגיש /, 2026-08-16) the phrase returns 42 articles, and insource:/ שב והדגיש / returns 11; both figures reproduced across three runs.

יוּסַר is a huf'al, and huf'al is not a choice

יוּסַר is the huf'al — the passive of hif'il — of הֵסִיר "to remove", from the hollow root ס־ו־ר. Its subject is the construct chain בַּרְזֶל־כְּבָלָיו, and a chain agrees with its head, not with its last word: the head is בַּרְזֶל, masculine singular, so the verb is masculine singular too — never mind the plural כְּבָלִים sitting next to it.

The point worth carrying away is structural. pu'al and huf'al are not free-standing binyanim you pick for flavour: pu'al is the passive of pi'el and huf'al is the passive of hif'il, mechanically. Choose hif'il הֵסִיר and its passive is already decided for you. And because they are derived voices, neither has an infinitive or an imperative — there is no *להוסר, and no way to command something to be removed.

Three more things in four words

• כְּבָלָיו is a plural bound noun with a third-person masculine suffix: כְּבָלִים "chains" plus "his/its". Speech would say הכבלים שלו.
• מֶנּוּ is the poetic short form of מִמֶּנּוּ, "from him / from it". It exists only in verse and in biblical quotation; write מִמֶּנּוּ.
• יָקוּם דּוֹר puts the verb before its subject. Hebrew's neutral order is subject-first, so verb-first here is a stylistic inversion — the same move Rachel makes with נָם חֶרְמוֹן הַסַּבָּא on the כנרת page.

The last line, עַיִן־בְּעַיִן יִרְאֶה אוֹר, is lifted from Isaiah 52:8, כִּי עַיִן בְּעַיִן יִרְאוּ בְּשׁוּב יְהוָה צִיּוֹן. Tchernichovsky keeps the phrase and changes the plural verb to a singular one to match his דּוֹר.

Common Mistakes

1. Parsing עוֹדֶנִּי as a verb. It has a verb's shape and none of a verb's morphology.

2. Reading שׂחקי with a shin. One dot, and it is a different root with a different meaning.

3. Giving huf'al an infinitive. There isn't one, and there is no imperative either.

4. Reading אאמין as a typo. The doubled alef is the grammar, not a slip.

5. Taking a fronted complement for a question. Word order does not ask questions in Hebrew.

Key Takeaways

• The title is אני מאמין; שחקי, שחקי is the incipit people use instead. Odessa, 1892; collected 1898; Tchernichovsky died in 1943, so the text is public domain.
• שַׂחֲקִי is ambiguous by design of the morphology: the pa'al feminine imperative ("laugh") and the pi'el feminine imperative ("play") are the same string. The masculine forms, שְׂחַק and שַׂחֵק, are not.
• זוּ is a relative particle, equivalent to אשר / ש־, not the demonstrative זוֹ. Exodus 15:13 has it too.
• עוֹדֶנִּי is an inflected particle. The family — אינו, ישנו, הנני, עודני — is literary at the עוד end and completely ordinary at the אינו end.
• יָשׁוּב יִפְרַח = "will bloom again". The modern descendants are שב ו־ and חזר ו־ plus a finite verb; the everyday word is just שוב.
• יוּסַר is huf'al, the automatic passive of hif'il הֵסִיר. pu'al and huf'al are derived, not chosen, and neither has an infinitive or an imperative.
• מֶנּוּ is poetic for מִמֶּנּוּ, כְּבָלָיו is a suffixed plural, and יָקוּם דּוֹר is verb-before-subject inversion.