In the first treatise of On the Genealogy of Morality, §5, Nietzsche proposes the following derivation of bonus (good):

I believe I may interpret the Latin bonus as "the warrior": assuing that I am correct in tracing bonus back to an older duonus (compare bellum = duellum = duen-lum, in which that duonus seems to me to be preserved). Bonus accordingly as man of strife, of division (duo), as man of war--one sees what it was about a man that constituted his "goodness" in ancient Rome. (translation by Clark and Swensen)

Obviously, a lot of important etymological scholarship by more level heads has happened since 1887. Is Nietzsche's proposed etymology for bonus plausible?

3 Answers 3


It is plausible that bonus is related to bellum/duellum

Nietzsche is correct about bonus coming from an older form starting with du-: not only is this consistent with known Latin sound changes, we apparently even have inscriptional evidence of forms starting with duon- and duen- in Old Latin. The reconstructed pronunciation of du- here is /dw/.

As Alex B. said in a comment, Michiel de Vaan's Etymological Dictionary of Latin (2008) says that "The best etymology for duellum so far has been proposed by Pinault 1987, who posits a dim. *duenelo- to bonus".

But modern linguists don't agree with the idea that bonus originally meant "the warrior"

However, Nietzsche's implication of an etymological connection between bellum/duellum and duo seems to be regarded less favorably by de Vaan, who says that the PIE source of bonus is unclear, and who says that the use of the form duellum with three syllables in pronunciation in later poets "may be either the result of folk etymology with duo, or a regular development".

And it seems that even though modern sources support a connection between bonus and bellum, they view the semantic connection as developing in the opposite direction: rather than an older word for warrior coming to mean good, de Vaan suggests that bonus (duenos) originally meant good, and the derivative bellum (duellum) was euphemistically used to describe war.


This etymology is not accepted by modern scholars, though the ultimate origin of "bonus" is contested. I cannot get de Vaan to scan properly, but here is Walde:

bonus „gut“, altlat. duonus, noch älter Duenos (nicht damit ablautend; vgl. Thurneysen KZ. XXXV, 204, Pedersen KZ. XXXVI, 91), vgl. auch bene, bellus: nach Darmesteter De conjug. latini verbi dare (1876), 26 ff. , Osthoff MU. IV, 370 ptc. auf -eno- zur Wurzel von ai. dúvaḥ n. „Gabe, Ehrerweisung“, durasyáti „ehrt, verehrt, erkennt an, belohnt“, duvasyú-ḥ, duvōyú-ḥ „verehrend, ehrerbietig“, wozu auch lat. beāre „beglücken, erquicken“, beātus „selig, glücklich“ ; hierher nach Wood Mod. Philology IV, 499 als idg. *du̯ei-to- vermutlich auch mnd. twīden „willfahren, gewähren, bewilligen, erhören“, mhd. zwīden ds., md. getwedic „zahm, willfährig“, vgl. auch as. tu(g)iđon „gewähren“, ags. tygđian, tīđian ds.; die Bed. „durch Gaben, durch Beschenkung ehren oder erfreuen“ (woraus die Bed. von ital. *du̯-ei̯ō, *du̯-enos sich leicht ergibt) ist der Hirt'schen (IF. XXI, 169ff.) Anknüpfung der Sippe an alat. duim usw. „ich möge geben“ (s. ; er führt idg. *dō-, *dōu̯- auf älteres *dou̯e- zurück) günstig. Sehr fraglich ist Zugehörigkeit von dautia, lautia, s. d.

Wenig ansprechend ist Fröhdes BB. IX, 111, Ficks I4, 457, II4, 150 (ebenso Prellwitz Gr. Wb. unter δύναμαι) Verbindung von bonus mit einer Wz. *dū-:*deu̯ā- „stark, fest“ in gr. δύναμαι.

  • 1
    Would it be possible to include a text transcription for those of us that have visual impairments that make it difficult to read text in an image? It would also make the text searchable. Nov 27, 2018 at 20:00
  • 1
    To expand on @AlexB.'s comment, de Vaan's explanation of the putative semantic link is this: "If *duenelo- meant 'quite good, quite brave', its use in the context of war (bella acta, bella gesta) could be understood as a euphemism, ultimately yielding a meaning 'action of valour, war' for the noun bellum". To me this sounds rather strained, especially given that de Vaan himself says of bellus (s.v. bonus) that it "was originally used to refer to women and children; it was applied to men only ironically".
    – TKR
    Nov 28, 2018 at 5:21
  • 1
    @AlexB. I understand the argument perfectly well -- it just strikes me as farfetched.
    – TKR
    Nov 28, 2018 at 19:45
  • 1
    @AlexB. Certainly. It's always hard to know how to evaluate such things.
    – TKR
    Nov 28, 2018 at 20:53
  • 2
    An English translation would be appreciated for those of us who don't speak German; I'd translate it myself but I'm afraid I'd miss a nuance.
    – Draconis
    Nov 29, 2018 at 18:29

The logic is sound: du = du, and duellum - du = ellum, much in the same way that du = du, and duonus - du = onus. Now in both cases where b = d, the results are the same: bonus and bellum.

Let's not ignore the consistency of the 'goodness of war' being implied in the etymology of both words.

reference: https://en.wikipedia.org/wiki/History_of_Latin

Initial *dw- (attested in Old Latin as du-) becomes b-, thus compensating for the dearth of words beginning with *b in PIE:

  • 3
    bellum/duellum "war" is cognate with Greek duh "agony", while bellus "beautiful" comes from an old diminutive of bonus/duenos. They're not related. And a single word's analogy isn't a solid argument: as Brian pointed out, bos "cow" ≠ duos "two".
    – Draconis
    Nov 28, 2018 at 2:32
  • 8
    This is a little silly: the unstated premise is that, "If a subset of letters behaves in a certain way for one word, it behaves in the same way for all words." Considering that optumus is a variant spelling of optimus, you would need to claim that all i's in Latin can be replaced by u's. It's a clear reductio ad absurdum.
    – brianpck
    Nov 28, 2018 at 2:57
  • 2
    @Draconis why do you think that bellum (OL duellum) and bonus (OL duenos) are not related?
    – Alex B.
    Nov 28, 2018 at 3:48
  • 2
    @AlexB. The etymology I've heard is that duellum is from PIE *d-h₂w "destroy" (> AGrk δύη "agony") while duenos is from PIE *d-w "revere" (> Lat beāre "bless"). But I wouldn't be surprised if de Vaan has something newer/more accurate.
    – Draconis
    Nov 28, 2018 at 4:37
  • 1
    Comments are not meant for extended discussion. Octavian, your conclusions were challenged, and the best response is to back up your answer by editing to add sources. For a free discussion environment, consider our chat. In case something seems out of order or there are questions, anything related to the functions of this site should be taken to our meta.
    – Joonas Ilmavirta
    Nov 29, 2018 at 9:25

Your Answer

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

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