In several languages, the word for "chapter" (a self-contained unitary text of a book) comes from the word for "head":

  • In Latin, "capitulum" (literally "small head") comes from caput (head). This word have derivatives in a lot of other languages, including English' "chapter".
  • In Greek, the word for chapter is "κεφάλαιο" ("cephalaio"), which comes from "κεφάλι" ("cepháli", head), which the english prefix "cephalo-" comes from.
  • In some Slavic languages, including Russian, chapter is "глава" ("glava"), which also means "head".
  • In Armenian, "գլուխ" ("glux") is both "chapter" and "head".
  • In Guarani, "akã" means both "chapter" and "head".
  • In Maori, "upoko" is both "chapter" and "head".

How is the word for chapter related to head in those languages. Which semantical derivation or metaphor was used in the formation of the words for chapter?

  • This question got 3 downvotes. Is there something wrong with the question?
    – Seninha
    Mar 28, 2018 at 12:05
  • 4
    I find linguists have an aversion to "why" in such cases - "how" a word came to mean something is enough and your question shows that you know how to research that already. I answered because I think there is a pretty good reason but note that it would be hard to find a definitive source for my answer, someone who knows for sure what the metaphor meant to the first person who used "head" this way. (Plenty of folk etymologies by overconfident Ancient Roman writers also spoil the broth.) I should clarify that in my answer. Mar 28, 2018 at 14:03
  • 1
    English "chapter" is a borrowing (via French) from Latin capitulum, not a cognate.
    – fdb
    Mar 28, 2018 at 16:40
  • I know that it's a borrowing. I'd chosen the wrong word. (But actually derivatives are cognates, aren't they?)
    – Seninha
    Mar 28, 2018 at 17:37
  • 1
    In everyday speech and in second-language instruction, people do tend to use "cognate" to mean a shared ancestor because recognizing those ancestors has such great practical use, but AFAIK as a technical term it refers to a genetic relationship to that ancestor, so no adoption. There might be some disagreement about it (I'm not a linguist) but here's @fdb's stance. Mar 28, 2018 at 18:32

3 Answers 3


It's useful in linguistics to distinguish between the 'why' and the 'how'.

  • The 'why' question is easy in the case of European languages because we know that in this part of the world, books and writing were first introduced in Greece and spread from there; in the medieval West this occurred almost exclusively via Latin. Thus we appear to have a direct line of transmission: Greek > Latin > other European languages (> non-European ones). Most Latin book-related vocabulary that wasn't directly borrowed from Greek was calqued from it, which also points to a Greek origin (an assumption that will be later proven wrong).
  • For French as well as all the other Romance languages, the answer to 'why' is 'it was borrowed from Latin', for English 'it was borrowed from French'; for all the other languages you list, seemingly apart from Greek, the answer is 'it was calqued from one of these other languages' (Slavic and Armenian from Greek, Guarani and Māori from Spanish/Portuguese and English respectively). Almost certainly the primary use of the term at first was to refer to chapters of the Scripture.

The most fundamental part of a calque is the borrowing of the metaphor. Finding out how this metaphor developed will give us the answer to the 'how' question, which will turn out to be much more interesting.

  • Happily, I've chanced upon a recent monograph by Charles E. Hill entitled The First Chapters, which treats precisely this topic. It explains that 1st-century CE Greek scrolls were divided into books using final colophons (e.g., Epicurus' On Nature, Book 2); however, smaller divisions don't appear to have been marked at this stage. From the 2nd century CE onwards, various devices start being used to indicate text divisions (including u̲nderlines, called paragraphoi), but neither the devices nor the divisions are standardised; it appears that the divisions used in literary works corresponded to our paragraphs, not chapters. Astonishingly enough, a 3~2nd century BC (!) Milan papyrus of Posidippus contains subject headings, although it's a manuscript of poetry and so it wouldn't be possible to talk about chapters as we understand them. It's not until the 3d century CE that we first encounter a head-related word referring to a textual division. The word for 'head' itself, κεφαλή, is never employed for this.

  • In contrast, the classical Latin caput is attested in reference to a division or section of a written work starting with Cicero (dated to around 80 BC); capitulum is its diminutive and appears in the same meaning later. We don't have any extended survivals of Roman literary papyri from that era, and scholars aren't clear on how these divisions were at first graphically indicated. There were hardly any division headings (rubrics), but division-initial letters were often somewhat larger than the rest and/or offset using hanging indent, thus protruding to the left (+- ekthesis). (My own guess is that the most basic device was breaking off the last line of the previous divison where it ended, so that section transitions were indicated by the empty space to the right margin.) Recall that the divisions themselves corresponded more or less to our paragraphs.

  • It's at this point (p. 27–8) that we come upon an especially interesting obervation:

    According to Butler, Latin speakers called these divisions of text capita [...] This, of course, exactly parallels one of the phenomena of textual division described earlier, usually called 'ekthesis', the projection of one or two letters into the lefthand margin. And, as I shall describe in more detail in Chapter 2, the resulting section was sometimes called by the Greek equivalent of the Latin diminutive capitula, namely κεφάλαιον.

    It goes on to say that ancient copies of Cicero speeches in particular may have ordinarily marked new divisions in this way, a practice perhaps originating with the author himself because it appears in all the suriviving manuscripts except one, which instead marks divisions with the letter K (for kaput). This alone indicates that such use of the word was well-established. (If the Google preview allows, make sure to read the following pages which contain some extended discussion.)

To sum up, according to Hill, the metaphor is actually Latin in origin, which would make the direction of borrowing surprising. The basis for the metaphor is the protruding section-initial letter that looks like a 'head' sticking out of the 'body' of the text. The first attested use of the diminutive capitulum in this meaning is with Tertullian, who died in AD 160 (see the TLL entry, column 351, line 38), and then primarily in relation to manuscripts of the Bible, but the dates seem to match up, with κεφάλαιον acquiring this meaning roughly a century later. Note that another diminutive, κεφαλίς, is used as equivalent to Latin volūmen 'roll', also in reference to the Bible.

Nevertheless, in Section 2.3.2 Hill goes on to explain the development of the Greek word as originating in the meaning 'the main point, summary (of a text or its part)'. Such summaries could be written in the margins (serving as headings) or abstracted into a list and affixed to the beginning or end of the scroll (serving as contents or index), in both cases amounting to chapter titles. The ultimate meaning 'chapter-section' is said to develop around the beginning of the 3d century CE, as already mentioned.

So as you can see, even one and the same author gives two different answers to the 'how' question for two different languages spoken in the same cultural space, indeed in a situation of habitual bilingualism. I'm not sure whether Hill explicitly reconciles the two explanations at any point, but it seems perfectly plausible that the 'head-letter protruding from the text-body' metaphor of Latin met the 'section summary abstracted into a heading' metaphor of Greek to produce the notion of 'chapter' that we know today.

Finally, I'd like to repeat the caveat that this 'how/why' explanation only applies to European languages (incl. Latin & Greek), and to those languages that received bookwriting together with textual terminology from the former. In these, the 'why' question is answered with 'because the terms have been borrowed or calqued.' There may very well be other languages that developed the same metaphor independently; in their case the 'how' is likely to be different, or with some luck it could be the same.

Addendum: I've now consulted Butler's super helpful article (check out that typography - worthy of the publication!). To, ehem, recapitulate his findings:

  • The practice of offsetting sections with hanging indents likely originated in Roman legal inscriptions around 100 BC. Its function was to mark a legal clause.
  • The name caput in reference to these sections probably appeared in parallel, in reference to the physically protruding letters of the indent. Alternatively, it could derive from the meaning 'the main point of an argument', as in Greek, and predate this inscriptional practice, but there's no evidence for this.
  • The first attested use in a Roman papyrus is from AD 27 and enough further examples are cited to suggest that the practice was wide-spread. An example of a non-documentary papyrus is a speech by the emperor Claudius (died AD 54). It was also widely employed on wax tablets as well as being adopted for all sorts of non-legal inscriptions. At that point its function was very much like that of our paragraph, a way of organising a written text.
  • This part is of particular interest to our inquiry:

    B. Maurenbrecher, who signs the [TLL] entry [for caput], connects the epigraphic convention he has in mind to the use of caput to designate the extreme end of a man-made object, considered either horizontally or vertically, as in the protruding end of a beam used for architectural support, or the end of a battering ram, or the head of a nail, or the capital of a column, or the two ends of a ruler or bolt of cloth, suggesting, therefore, that caput, in its legal use, designated the protruding “head” (presumably the beginning) of a section, set off by a few letters from the main block of the text by a convention of legal mise-en-page.

  • (I was right about basic line-breaks serving as section breaks :-)



We can't know exactly which quality led so many languages to independently develop or borrow the metaphor — etymological dictionaries rarely speculate on the "why" — but here are my thoughts.

There are at least three qualities of "head" that have a possible metaphor with "chapter".

  1. Head = (most) important. So natural that examples are hardly needed.

  2. Head = first. "Headwaters", "headline", "head off" (get there first), "ahead" (first), etc.
    For some cross-linguistic confirmation from an unrelated language, consider e.g. Hebrew Rosh Hashanah lit. "head of the year", translated "New Year's Day". (The root of rosh incidentally appears in many words, including bereshit usually translated "in the beginning").

  3. Head = orientation. "Ahead" (in the direction faced), "head somewhere", "head fake", etc. Perhaps by connection with face (which direction are you facing?).

Let's investigate using the OED. Their etymology is a good example of how we usually ignore the "why" while charting the "how"; nothing here identifies the link used in the metaphor:

Etymology: A later syncopated form of chapiter n., < Old French chapitre, earlier chapitle < Latin capitulum. diminutive of caput head, used, in ancient Latin, in the senses ‘little head, head of a plant, capital of a column’, and later, those of ‘head-dress of women, chapter of a book, section of a law’. The form chapter appears in Scots in 14th cent., but in English is rare before the 16th; chapiter survived beside it till the middle of the 17th, and is still occasional in the sense ‘capital of a column’.

But I think the definitions themselves give us clues that (1) above is the relevant link:

1. a. A main division or section of a book (whether the latter is an entire literary work, or one of the divisions or parts of a large work). Esp. used of the main divisions of the books of the Bible.

There are examples from as early as c. 1000, but this is the earliest one that's unambiguous to me:

c1386   Chaucer Nun's Priest's Tale 245   In the same book..Right in the nexte chapitre after this.

The ambiguity of the earlier examples arises from overlap with this definition:

2. fig. Head, heading, subject, category. (Usually preceded by on, upon.) arch.

A seemingly relevant example:

1681   W. Temple Mem. iii, in Wks. (1731) I. 342   Upon which Chapter I said a good deal.

And one more whose examples are much later but possibly related to the underlying mechanism:

5. c. A branch of an organization or society, esp. of a college fraternity. U.S.

Based on these, we could propose a semantic development along these lines:

(broadening) important thing
(narrowing) among some group of collected items, e.g. people, the representative or chief
(sideways/synecdoche) something having been organized, a main division
(narrowing) a main section of a book
(broadening) any section of a book

The paradox is how a word that means "head" came to mean "subdivision". We have to switch our point of view: Looking at a book, the chapters are subdivisions. But looking at the mass of words, each chapter heads a group of them. The chapters of an organization are branches of the whole, but in terms of the members a chapter collects them under a single head(ing). And again, the world has any number of things that are grouped into the "categories" of definition 2. And not just any sections, but the "main" sections of definition 1c. To start with, anyway.

So a chapter is a "head" not looking from the top down, but from the bottom up.

  • Wouldn't "in the beginning" be bareshit (with a pathakh under the bet)? So bereshit should be "in A beginning". Mar 28, 2018 at 10:32
  • 2
    @Wilson It would be a qamats (resh can't be doubled as the def. art. requires and the patakh lengthens to compensate). But interestingly, it is actually a sheva in Gen 1:1 as you guess, indicating no def. art. The Jewish Publication Society translation reads a participle or a construct noun (with bara as the head), and either way renders it "When God began to create". This article argues for it as an indefinite noun and renders "in a former state". Others argue textual corruption. I just went with "usually translated" above :) Mar 28, 2018 at 13:55

In Dutch we have chapter="hoofdstuk" (head piece, literally). "hoofd" (head) can both mean literal head (body part) and figurative (leader/boss), as English head can also mean. The semantics here is "hoofd" as main/governing part of the body --> governing/most important person in general. So the "main parts" of the book are called hoofd-stuk. In Westerlauwer Frisian the same theme can be seen; here we have two words "holle" (for the body part) and "haad" (for the boss) and a chapter is "haadstik", (not "*holstik" or some such). Capital (also caput = head) letters are also "hoofdletter/haadletter" resp. So I'd look into the head=boss semantics in the languages you mentioned as well.

  • So until now we have 3 proposed semantic developments: Chapter is named thus because it's the main (head) parts of a book (your proposal); chapter is named after the book's orientation (Luke's proposal); and it is named after the heading, a short paragraph which identifies a chapter (my proposal). I still haven't found a etymological dictionary explaining the correct semantic evolution.
    – Seninha
    Mar 31, 2018 at 20:40
  • This explanation doesn’t work for Dutch kopje, used for the heading of smaller sections and sections themselves.
    – Keelan
    Apr 14, 2022 at 4:33
  • You demand that there be only one reason for a continent-wide tendency? Nothing is that simple. And there are plenty more reasons for the HEAD metaphor; as we've known for a long time, the human body and its parts, being the only thing all humans have in common, are the common source of almost all metaphors, and is certainly implicated here. What, after all, is the measure of all things?
    – jlawler
    Apr 19, 2022 at 15:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.