Obviously, I would not have embarked on such a long manhunt if I did not have my idea on the origin of the troublesome word. It will probably end up in the dustbin (also known as ash heap) of etymology, but there it will come to rest in good company.
It seems that the Goths and the Old Scandinavians preserved the early stage of secularizing Mannus’s name (Mannus, let me repeat, was, according to Tacitus, the supreme deity of the Early Germanic peoples). Gothic had the noun gaman (neuter; ga– is a collective prefix) “fellowship” and (!) “partner.” Old Norse man, neuter or feminine (!), meant “bondsman” and “maid; concubine.” Old Norse had lost all prefixes before the earliest texts in that language were recorded, so that man in it can be an exact analog of Gothic gaman. I suggest that in the beginning gaman referred to a group of Mannus’s worshipers (by the way, the most ancient form of Mannus had one n, and that is why in the earlier posts I sometimes enclosed the second n in parentheses; the origin of that additional n has been discussed at length but need not delay us here).
That the same word sometimes designates a group and an individual will not surprise anyone. I mentioned this fact in dealing with the noun god (originally, only the neuter plural—Old Norse guð “gods”—existed) and in my brief discussion of the origin of wife. Engl. youth “young people” and youth “a young person” is not an isolated case, and in everyday speech we hardly notice the ambiguity. From a long list of such examples collected by historical linguists I can mention Old Engl. leod “people” (a recognizable cognate of Modern German Leute “people”) and “prince.” German Stute “mare” corresponds to Engl. stud “a group of horses kept for breeding; a male horse belonging to such a group”); Romanian feméi, originally “family,” now means “woman.” Latin manes “deified ghosts of the dead” later acquired the meaning “corpse.”
Nor should the neuter, when applied to a human being, look like an anomaly. German Mensch “human being” is masculine (der Mensch), but at one time it had a parallel neuter form (das Mensch; their coexistence is still recognized by the modern language); in one of the Old Scandinavian dialects, þiuþ “man” was neuter (þ has the value of Engl. th in thin). Grammar “overrides” semantics and sometimes produces absurd results. For instance, in German, in which the diminutive suffix –chen always makes a noun neuter, Mädchen “girl” (a cognate of Engl. maid) is, naturally, also neuter. To be sure, every form of this type should be explained on an individual basis, but the need for special pleading does not invalidate the fact that the name for a single human being can end up as a neuter noun.
In medieval German, man had numerous meanings, and some of them, such as “man, male, son; brave warrior; lover, fiancé,” are easy to understand, but in courtly poetry “vassal” predominated; compare Engl. all the king’s men. In English chess, every piece except the king can be called man; hence chessman. Although man early acquired the sense “a male,” the German pronoun man (e.g. man sagt “one says”) and its use in jemand “somebody, someone” and niemand “nobody, no one” (-d is a later accretion in both words) are witnesses to the ancient “gender neutrality” of those forms. Engl. woman, a compound of wif and man, meant something like “a person called wif.” English would, most probably, have retained the pronoun man in the function similar to that of German man, but under the influence of French on, as in on parle “one speaks,” it was ousted by the similar-sounding one, which is amusing, because French on goes back to Latin homo “man.”
Given such facts, I suggest that the development was from “fellowship in Mannus” to “a fellow in Mannus” (“partner,” masculine, female, or neuter by default), further to “human being” and to “a person of low status,” first in relation to the deity, then to the lord: “vassal, slave, concubine, etc.” If I am right, man is a relatively late word, devoid of cognates in the rest of Indo-European almost by definition. “Late” is a loose concept: a runic inscription going back to the beginning of the Common Era already had Man(n)R. Yet nineteen hundred or two thousand years would indeed be a relatively short period in the entire history of Indo-European. Our word’s “young age” may be the reason why its grammatical forms behaved so erratically: man- was declined according to several types. But such fluctuations are not too rare even today. For instance, Modern German has three plural forms of Mann: Männer, Mannen, and Mann (each has its sphere of application).
It will be only fair to say that long before me two great scholars—Jacob Grimm and Friedrich Kluge—came to the results similar to mine, but Grimm did not go beyond stating that, although man at one time meant “servant” (this being its main meaning!), one should not think that Old Germanic people (Germanen) were slaves, while Kluge, who reconstructed the original sense of man as “the progeny of Mannus,” gave up his idea, for he could not resist the temptation of finding an ancient etymon of the Germanic noun and invented an improbable hybrid of man and guma, mentioned in the first part of this series.
We now have to search for the etymology of the god’s name, but here only a vague guess is possible. Of prime importance is the fact that m-n conveys nearly the same idea in various languages: Korean myång, Chinese manu, Austronesian muani, and many others like them. Among the meanings of this sound complex we find “man, boy; a phallic deity; herdsman; warrior; woman; people” and “kin,” some of them familiar from the material of Germanic. Wilhelm Oehl, to whom I have often referred in my recent posts, considered man a universal baby word. In Indo-European, the syllable man is often connected with the idea of evil spirits, manias, and madness. Manus probably arose in human conscience as a frightening creature; this is the most common scenario in the history of religion. Like some other gods, with time he acquired benevolent features, but, in principle, the ruthless deities of old sent physical and mental diseases to their worshipers, who propitiated them with sacrifices. This subject has also been developed in the series of posts on the origin of god. The syllable man may be a baby word. If so, it first designated a bogeyman with whom to scare little children.
My tentative progression is from a baby word (like boo) to a materialized demon, perhaps producing that terrifying sound, to a supreme awe-inspiring divinity, and to the main god of the ancient Germanic pantheon. People were believed to be his servants, and their group was galled gaman. Any member of the gaman was a “man,” that is, a human being, originally a worshiper of either sex, a servant of the great deity. Much later the meaning was specialized to “a male,” but we still have manholes, even if though they have been renamed sewers, manhunt, manpower, and of course mankind.
Image Credit: (1) “Mannus tre söner, gravyr av Justus Peterson efter en bild av Carl Larsson” by (2) “Page 5 of Denslow’s Humpty Dumpty” by theornamentalist. Public Domain via Wikimedia Commons. (3) “Chess Play” by Remco Wighman. CC BY ND 2.0 via Flickr.
Anatoly Liberman is the author of 
Is that the same “gaman” that also meant pleasure, enjoyment, togetherness and is the root of contemporary english “game”?