Spinoza's Compendium of Hebrew Grammar (1677, posthumous, unfinished) contains a claim that scholars have been misreading for centuries. He says that all Hebrew words, except a few particles, are nouns. The standard scholarly reaction is that this is either a metaphysical imposition (projecting his monistic ontology onto grammar) or a terminological trick (defining "noun" so broadly it's vacuous). Both reactions wrongly import Greek and Latin grammatical categories and then treat those categories as the neutral baseline.

Spinoza's Claim: Hebrew's All Nouns

From Chapter 5 of the Compendium (Bloom translation, 1962):

"By a noun I understand a word by which we signify or indicate something that is understood. However, among things that are understood there can be either things and attributes of things, modes and relationships, or actions, and modes and relationships of actions."

And:

"For all Hebrew words, except for a few interjections and conjunctions and one or two particles, have the force and properties of nouns. Because the grammarians did not understand this they considered many words to be irregular which according to the usage of language are most regular."

The word "noun" here is nomen. It means "name." Spinoza is saying: almost every Hebrew word is a name for something understood. This includes names for actions, names for relationships, names for attributes. His taxonomy of intelligible content explicitly includes actions and modes of actions alongside things and attributes.

The Vacuousness Objection

The obvious objection is: if "noun" covers actions as well as things, then the claim that "all words are nouns" is trivially true and does no work. Any content word names something intelligible; so what?

But this objection assumes that a useful grammar must draw a hard categorical line between nouns and verbs, and that Spinoza's refusal to draw it is therefore vacuous. That assumption is embedded in the Greek grammatical tradition; it is not a fact about Hebrew.

Semitic Roots

In Hebrew (and Arabic, Akkadian, and other Semitic languages), words are generated from consonantal roots—typically trilateral—by applying vowel patterns and affixes. The root כ-ת-ב generates katav (he wrote), kotev (one who writes), ktav (writing/script), mikhtav (letter), katvan (scribbler). The morphological operation is the same in every case: take the root, apply a pattern that describes the relation of the concept to the thing you are describing. For example, mikhtav is something that is made-written, a letter, much like the Arabic mameluke is someone who is made-owned, a slave. Whether the output functions as what a Greek grammarian would call a "noun" or a "verb" depends on which pattern you applied, not on some fundamentally different generative process.

This is not how Greek or Latin works. In those languages, nouns and verbs belong to largely separate inflectional systems (though they do have participles). Nouns decline for case and number; verbs conjugate for tense, aspect, mood, and person. A Greek speaker can usually tell from a word's form alone which category it belongs to. The noun/verb distinction corresponds to a real difference in morphological machinery.

In Hebrew, it doesn't. The grammarians who insisted on the distinction—both the rabbinical grammarians working in the Arabic tradition and the Christian Hebraists working from Latin—were forcing Hebrew into a framework designed for languages with a different structure. The result, as Spinoza observed, was that regular Hebrew forms got classified as irregular, because they didn't respect a boundary the language doesn't draw.

Where the noun/verb distinction comes from

The Arabic grammatical tradition, which the medieval rabbinical Hebrew grammarians adopted wholesale, classifies words into three categories: ism (noun/name), fi'l (verb/action), and ḥarf (particle). Scholars have long noted the parallel between this trichotomy and Aristotle's division of speech into onoma (name), rhema (verb/predicate), and sundesmos (connective); Syriac scholars were important intermediaries in transmitting Greek linguistic thought to Arabic, though the degree of direct dependence remains debated. ^[1] The classification reached Hebrew grammar through two independent routes: Greek → Latin → Christian Hebraists, and Greek → Arabic → rabbinical grammarians. Both paths originate in Greek philosophy.

Judah ben David Hayyuj (c. 945–1000), the founder of scientific Hebrew grammar, applied Arabic grammatical theory to Hebrew, including the ism/fi'l/ḥarf trichotomy and the principle that all Hebrew roots are trilateral. ^[2] His technical terms were translations of Arabic grammatical terms. Jonah ibn Janah (c. 990–1055) extended this work, producing the first complete Hebrew grammar and drawing explicitly from the Arabic grammatical works of Sibawayh and al-Mubarrad. ^[3] When Spinoza complained that "the grammarians" misunderstood Hebrew, this is the tradition he was arguing against.

Aristotle's noun/verb distinction is not just a grammatical observation. It reflects his substance/predication ontology. The world consists of substances (things that exist independently) and predicates (things said about substances). A noun names a substance; a verb predicates something of it. The sentence "Socrates runs" has the structure: substance + predication. The grammar encodes the metaphysics.

Dramas and Graphs

Greek and similar languages have different pools of words for filling the grammatical roles of noun and verb. Hebrew has one pool of roots that supplies words for both roles, depending on the pattern applied. These aren't just two different ways of doing the same thing. They reflect different structural priorities.

The Indo-European system is built around assembling a scene: placing distinct actors into relationships with distinct actions. You need different building blocks for the actors and the actions because they play different structural roles in the scene. Who did what to whom, when, in what manner. Case endings on nouns tell you the role; verb conjugation tells you the temporal and modal frame. The grammar presupposes that the actor/action distinction is primitive.

The Semitic system works differently. Each root is a node in a flat graph of intelligibles. The graph doesn't recurse; roots refer to intelligible things, not to relations between other roots. And it doesn't privilege any type of node over any other, which is why the morphological system treats them all with the same machinery. It does not start by assigning one word the role of "the thing" and another the role of "what the thing does."

A sentence picks out some nodes from this graph, and casts them into some definite relation to each other. Their arrangement and patterns of modification describe the way in which these intelligibles are related: process, agent, result, instrument, quality, location.

When you take the Greek-derived framework and impose it on Hebrew, you're asking a flat graph of intelligibles to behave like a scene-assembly system. The spurious irregularities Spinoza complained about are projections of the friction from this mismatch.

I'm not projecting, you're projecting!

The standard scholarly line is that Spinoza projected his philosophical commitments onto his grammar; that his monism (one substance, everything else is modes) motivated his claim that Hebrew has one part of speech with subcategories rather than two fundamentally different parts of speech. Harvey (2002) argues that the Compendium's linguistic categories parallel the conceptual categories of the Ethics. ^[4] Rozenberg (2025) goes further, claiming Spinoza "project[ed] the characteristics of Latin onto Hebrew" and thereby "neglected the dynamism of Hebrew." ^[5] Stracenski provides a more sympathetic reading but still frames the question as whether the Compendium serves the Ethics' metaphysics or the Tractatus' hermeneutics. ^[6]

This gets the direction of explanation backwards, or at least sideways. Spinoza was reading a Semitic language and describing how it actually generates words. The fact that his description aligns with his metaphysics may reflect a common cause: both the grammar and the metaphysics are what you get when you don't take the Aristotelian actor/action distinction as a primitive. Spinoza rejected Aristotle's substance/predicate ontology in the Ethics; he also noticed that Aristotle's noun/verb grammar didn't fit Hebrew.

1. Aristotle divides lexis into onoma, rhema, and sundesmos in Poetics 1456b–1457a. Farina documents how this tripartite scheme reached Arabic grammar via Syriac translations of Aristotle's Organon, with Syriac Christians serving as intermediaries between Greek and Arabic linguistic thought. See Margherita Farina, "The interactions between the Syriac, Arabic and Greek traditions," International Encyclopedia of Language and Linguistics, Third Edition, 2025. The question of whether Sibawayh's ism/fi'l/ḥarf directly derives from Aristotle or represents independent development remains actively debated; the structural parallels are clear even if the exact transmission pathway is contested. ↩︎

2. On Hayyuj's application of Arabic grammar to Hebrew and his establishment of the trilateral root principle, see the Jewish Encyclopedia entry on "Root". His Wikipedia biography notes that "the technical terms still employed in current Hebrew grammars are most of them simply translations of the Arabic terms employed by Hayyuj." ↩︎

3. Ibn Janah's Kitab al-Luma was the first complete Hebrew grammar. It drew from Arabic grammatical works including those of Sibawayh and al-Mubarrad. See also the Jewish Virtual Library entry on Hebrew linguistic literature. ↩︎

4. Warren Zev Harvey, "Spinoza's Metaphysical Hebraism," in Heidi M. Ravven and Lenn E. Goodman, eds., Jewish Themes in Spinoza's Philosophy (Albany: SUNY Press, 2002), 107–114. ↩︎

5. Jacques J. Rozenberg, "Spinoza's Compendium: Between Hebrew and Latin Grammars of the Middle Ages and the Renaissance, Verbs versus Nouns," International Philosophical Quarterly, online first, October 26, 2025, DOI: 10.5840/ipq20251024258. ↩︎

6. Inja Stracenski, "Spinoza's Compendium of the Grammar of the Hebrew Language," Parrhesia 32. Stracenski notes the divide between historical approaches (Klijnsmit, placing Spinoza within Jewish grammatical tradition) and philosophical approaches (Harvey, connecting the Compendium to the Ethics). See also Guadalupe González Diéguez's companion chapter in A Companion to Spinoza (Wiley, 2021) and Steven Nadler, "Aliquid remanet: What Are We to Do with Spinoza's Compendium of Hebrew Grammar?" Journal of the History of Philosophy 56, no. 1 (2018): 155–167. ↩︎