Safire’s latest “On Language” column [archived] is a fairly pointless trudge through various phrases used to describe the current conflict in Iraq; what caught my attention was the end of the first paragraph: “…others who see it as more political than religious call it an insurgency or an internecine (in-ter-NEE-sin) struggle.” (I have added the italics from the printed article, and it strikes me as bizarre in the extreme that the Times doesn’t bother to carry over the italics in the online version, since they are necessary in separating words presented as words from words used in the normal way as referents; if I were Safire, I would force them to remedy this. I note that the Houston Chronicle manages to preserve the italics when they reprint the column.)
It truly surprises me that Safire passed up a chance to expatiate upon the word internecine; he could have written an entire column on that alone. I’ll start by quoting Fowler‘s entry (first edition, of course; there’s no point reading diluted Fowler):
internecine has suffered an odd fate; being mainly a literary or educated man’s word, it is yet neither pronounced in the scholarly way nor allowed its Latin meaning. It should be called ĭnter’nĭsĭn, & is called ĭnternē’sīn; see False quantity. And the sense has had the Kilkenny-cat notion imported into it because mutuality is the idea conveyed by inter- in English; the Latin word meant merely of or to extermination (cf. intereo perish, intercido slay, interimo destroy) without implying that of both parties. The imported notion, however, is what gives the word its only value, since there are plenty of substitutes for it in its true sense—destructive, slaughterous, murderous, bloody, sanguinary, mortal, & so forth. The scholar may therefore use or abstain from the word as he chooses, but it will be vain for him to attempt correcting other people’s conception of the meaning.
The American Heritage Dictionary shares his sensible approach and adds some interesting detail about where exactly the change in meaning came from:
When is a mistake not a mistake? In language at least, the answer to this question is “When everyone adopts it,” and on rare occasions, “When it’s in the dictionary.” The word internecine presents a case in point. Today, it usually has the meaning “relating to internal struggle,” but in its first recorded use in English, in 1663, it meant “fought to the death.” How it got from one sense to another is an interesting story in the history of English. The Latin source of the word, spelled both internecīnus and internecīvus, meant “fought to the death, murderous.” It is a derivative of the verb necāre, “to kill.” The prefix inter– was here used not in the usual sense “between, mutual” but rather as an intensifier meaning “all the way, to the death.” This piece of knowledge was unknown to Samuel Johnson, however, when he was working on his great dictionary in the 18th century. He included internecine in his dictionary but misunderstood the prefix and defined the word as “endeavoring mutual destruction.” Johnson was not taken to task for this error. On the contrary, his dictionary was so popular and considered so authoritative that this error became widely adopted as correct usage. The error was further compounded when internecine acquired the sense “relating to internal struggle.” This story thus illustrates how dictionaries are often viewed as providing norms and how the ultimate arbiter in language, even for the dictionary itself, is popular usage.
I have to quarrel, though, with the sentence about “the Latin source of the word”; surely internecīnus and internecīvus are two different words, not two spellings of the same word.
And the OED explains the source of the “false quantity”:
[App. first used as a rendering of L. internecīnum bellum, in Butler’s Hudibras (to which also is due the unetymological pronunciation, instead of in’ternecine). On this authority entered by Johnson in his Dictionary, with an incorrect explanation, due to association with words like interchange, intercommunion, etc. in which inter- has the force of ‘mutual’, ‘each other’. From J. the word has come into later dictionaries and 19th c. use, generally in the Johnsonian sense.]
Butler’s line is:
1663 BUTLER Hud. I. i. 774 Th’ Ægyptians worshipp’d Dogs, and for Their Faith made internecine war.
In the 1674 edition he changed it, for whatever reason, to “fierce and zealous war.” Too late, though, for the pronunciation of the word.
I feel compelled to add that Garner gets it completely wrong in his Modern American Usage, starting his entry “As originally used in English, and as recorded by Samuel Johnson in 1755, internecine means ‘mutually deadly; destructive of both parties.'” I’ll mention also that Safire oddly chooses the OED’s pronunciation over the one preferred in American dictionaries (inter-NESS-een).
Update (Mar. 2026). The OED revised its entry in 2015; the senses are now “Characterized by great loss of life; deadly, very destructive, devastating. Obsolete.” (1642 “How bloudy were the persecutions raised against the Waldenses.., what internecine wars were stirred up against the Hussites in Bohemia?”), “In which either side aims to kill or destroy the other. Also in weakened sense: mutually antagonistic, bitter” (1755 “Internecine, endeavouring mutual destruction.” S. Johnson, A Dictionary of the English Language), “Of an enemy, enmity, hatred, etc.: mortal, deadly, bitter; (also) implacable, irreconcilable. Now rare.” (1692 “That internecine hatred which concludes in nothing but blood”), and the newly added sense 3:
Of conflict, rivalry, etc.: that takes place within a group, society, or organization; internal, civil, domestic.
In early quots. perhaps merely a contextual use of sense 1b.1834 The flagrant repugnance, political and personal, amongst the leaders of the liberal party,—the internecine war between Lords Brougham and Durham.
Standard 27 October1846 France..though powerful enough to have defied, if not to have mastered its continental neighbours, was still so difficult to preserve against internecine dangers, and to defend against itself.
C. F. Henningsen, E. Europe & Emperor Nicholas vol. III. iv. 155
[…]1977 Everyone..helped one another, at the same time maintaining a friendly internecine rivalry.
J. Draper in S. Kostof, Architect 2231991 Internecine wars among the Cossack hetmans, uprisings against the tsar, and pestilence and starvation devastated the land.
New York Review of Books 22 October 58/32013 Internecine warfare..has blighted Conservative political and fundraising activity in Thirsk & Malton for the past several years, creating schisms and unpleasantness between friends.
Northern Echo (Nexis) 18 December
The new etymology:
< classical Latin internecīnus, variant of internecīvus fought to the death, devastating, murderous (see internecive adj.).
Notes
The later semantic development of the word, starting with Johnson (see quot. 1755 at sense 1b), reflects reanalysis as if < inter- prefix + classical Latin nec-, nex violent death (see necation n.) + ‑ine suffix¹. Johnson cites Butler (compare quot. 1663 at sense 1a), who uses the word in its classical Latin sense, as the earliest authority for the English adjective. In sense 3 perhaps also influenced by a reanalysis of the main element as related to Latin nexus the action of binding, a bond, tie, a type of legal obligation, a combination, connected group.
According to Lewis & Short internecīnus and internecīvus are truly interchangeable – both meaning “deadly, murderous, destructive”. The first form seems to occur more frequently in classical Latin than the latter. I’m not enough of a Latin scholar to declare authoritatively whether or not the two forms originally had different nuances of meaning.
Also, in contradiction to AHD, Lewis & Short asserts that the third meaning of ‘inter-‘ in compound forms is not an intensive, but rather ‘under, down, to the bottom’. This seems to be how the form shows up in Latin words about death & killing. The word ‘inter’ is related to G. unter, E. under, Goth. undar, Sansk. antar.
I had thought “inter” in the sense komfo,amonan suggests is present in “interment,” a synonym for burial. But no, our “inter” there is late medieval Latin for…wait for it…burial.
(It’s also in Inter Milan, a team that routinely gets buried. Heh.)
A word sililiarly used which seems wrong but isn’t is “intestine”.
Well, at the risk of being obvious; the verb ‘inter’ is in + terra + are… perhaps this has influenced the development of the prefix ‘inter’ in certain contexts. But as a preposition, as in intereo (inter-ire, to go between/under) it is clearly euphemistic; compare ob-ire, to go down or die, per-ire, to die, etc.
I am distressed to discover that I have been pronouncing this word the wrong way (the “scholarly” way, according to Fowler) my entire life. Shocking! Well, now I have a pedant’s justification for it.
I can report that the abovementioned original sense of “Kilkenny cat” is no longer common knowledge in Ireland. https://www.joe.ie/sport/watch-there-was-a-fairly-odd-reference-to-kilkenny-hurling-on-the-569763
Ah, so this is another word where the -e is a lie.
I say “inter-NESS-in” myself, making the -e even more of a lie.
The OED revised its entry in 2015; see Update.
Forcellini’s dictionary, the ultimate source (via Freund’s Latin-German dictionary) of both Lewis & Short and Georges, treats them explicitly as spelling variants. Georges decided not to list internecinus at all, only internecivus (with the variant internicivus in Cicero). In medieval handwriting u/v and n could look very similar (basically, two vertical fat strokes with a thin stroke somewhere connecting them) and apparently Forcellini and Georges thought that the form internecinus was the result of a misreading of interneciuus.
The British pronunciation given by LPD and CEPD is /-ˈniːsaɪn/, pronunciations without the diphthong in the final syllable are American-only.
I like the OED’s new formulation using “variant.”
The Oxford Latin Dictionary lists internecīnus as a variant occurring once (Cicero, Philippics 14.7, as “iniustae belli internecini notae”.) The main entry in internecīuus, “also internicīuus.”
Johnson was not taken to task for this error. On the contrary, his dictionary was so popular and considered so authoritative that this error became widely adopted as correct usage.
This is just another example of Eddyshaw’s Law: “No Latin word means the same as any English word derived from it.”
Lingua latina non penis canina, as the Russians say.
“O civile, si ergo fortibus es in ero…” etc.
A word sililiarly used which seems wrong but isn’t is “intestine”.
Johnson sv “intestine”:
——
3. Domestick, not foreign. I know not whether the word be properly used in the following example of Shakespeare : perhaps for mortal and intestine should be read mortal internecine.
——
(The quote is by Duke Solinus in A Comedy of Errors, not a rude mechanical in Pyramus and Thisbe.)
It seems like the wrong word to me. It suggests that some manuscript(s) have internecinus and others don’t.
But internecinus apparently doesn’t occur in any manuscript. It’s a conjecture of Ferrarius in 1542 on Cicero, Philippics 14.7.
Mueller’s Teubner edition (1893) and the first edition of Clark’s OCT (1900) both adopt Ferrarius’ conjecture in their text, but they give it wrongly as “inustae belli internecivi notae”–inadvertently, or because they thought that was what Ferrarius ought to have written? But Clark’s 2nd edition (1918) gets it right as “inustae belli internecini notae”.
The manuscripts have:
in iusta evelli inter necti nota s
iniusta evelli internecuno te otv
iniusta evelluntur nec uno te h
iniuste belli interiectionis b
Good point. I liked “variant” because I took it for granted that both forms existed in the language, but if not, then you’re right, it’s not a good word.
Scary.