In the Indo-European languages, most words for “mother,” “father,” “son,” and “daughter” are very old—most (rather than all), because some have been replaced by their rivals. Thus, Latin filia “daughter” is the feminine of filius “son,” and filius has nothing to do with son, which is indeed ancient. The words for “father” and “mother” are especially stable, though even they may disappear under the attack of baby words. In English, dad ~daddy coexists with father, but Gothic, a fourth-century Germanic language, had only atta for “male parent,” seemingly a baby word, like Russian tyatya and Hittite tata, among others. Substitute d for t, and tyata will be almost indistinguishable from daddy.
The puzzling thing is that atta, in exactly this form, has been recorded in many languages, not only Indo-European. Sometimes this complex means “father,” sometimes “mother,” and sometimes “sister.” Many people will recognize the title Atatürk “father of the Turks” and Attila, the name of the fifth-century ruler of the Huns. Although Hunnish, he became known under the Gothic name that meant “Daddy.” Like other dictators with unlimited authority, he wanted to be known as the great parent of his tribe (or nation). The part –ila in his name is a diminutive suffix, related to –ula in Latin ursula “little she-bear,” here probably a sign of endearment (something like “our beloved father”). It is not related to but has the same meaning as the Modern English suffix –ling, which will reemerge in this story.
The existence of such words throws a curious light on the problem of cognates, or congeners. Why did so many languages have nearly the same name for a close relative? Those words are hardly related across language borders. It is even more unlikely that we are dealing with so-called migratory words or borrowings. Even though all infants babble in the same way, it is amazing how many identical baby words for “father” exist all over the world. Surely, a proto-word meaning “father” did not exist half a million years ago.
In the past, forming a family in Europe was a strongly regularized procedure, and it is still such in many parts of the world. Hence a multitude of words for every member of a large family: “maternal uncle” versus “paternal uncle,” “married woman in relation to her mother-in law” versus “married woman in relation to her father-in law,” “two men married to sisters”—people needed a term for everybody. English has lost all the complex terms of the past. It is enough to add “in-law,” to produce the entire panoply of parents, brothers, and sisters. Only cousin, a borrowed word, is still with us. A study of kin terms is a flourishing branch of etymology and of anthropological linguistics.
Child is not a kin term, and one cannot expect uniformity in the coining of this word across the globe or even in related languages. Some such names are clever and simple. The German for “child” is Kind, recognizable to English speakers from Kindergarten. The word must have arisen with the sense “belonging to or born to one’s kin.” Engl. kind, as in one of a kind and the indispensable kind of is an obvious cognate of German Kind. In the past, the English noun kind meant “birth, descent” (compare kith and kin). Likewise, the related adjective kind meant “natural” or “native” (as in kindly Scot). A close semantic neighbor of German Kind is Scots bairn “one born,” derived from the verb bear “to give birth.” Child has an ascertainable etymology. Its root shows in Gothic and means “womb.” Thus, child is the fruit of the womb. This makes perfect sense.
Equally transparent, but not at first sight, is the Slavic word for “child.” I’ll cite the plural form deti “children.” Deti is related (akin!) to Gothic daddjan “to suckle.” The Slavic child is thus a suckling, but only from an etymological point of view, for the one recognizable Russian cognate of deti is doit’ “to milk”; no native speaker of Slavic will dream of connecting them. However, once we notice such words, we are no longer surprised that Latin filius “son,” mentioned above, and femina “woman” share the root with deti (that is, if we know that Latin f- corresponds to Germanic b-, both from the reconstructed Indo-European consonant bh).
One’s child is a dear or a nuisance, as the case may be, and that is why various words have been invented to express people’s love or mild disapproval of a son and a daughter. Now we are, fortunately, unisex, but in the past, when little girls were made of sugar and spice, and all that’s nice, while little boys were constituted of frogs and snails, and puppy-dogs’ tails, a few amusing examples of “stigmatizing” male children turn up. Perhaps the most surprising of them is urchin. The word came to English from French and meant “hedgehog.” The development of senses must have been from “hedgehog” to “bristling creature,” to “contentious, mischievous youngster,” and occasionally “ragamuffin.” Those studying historical semantics distinguish between the amelioration and the deterioration of meaning. I am not sure whether the progress from “hedgehog” to “mischievous, ragged youngster” is improvement, regress, or neither. From the human point of view, it is certainly better to be a boy than a prickly beast, the more so as hedgehogs do not live in the New World (the closest approximation is the porcupine). The hedgehog’s point of view would be hard to assess.
The word imp moved almost in the opposite direction. The earliest recorded meaning of imp is “young shoot, sapling.” The corresponding verb (to imp) meant “to engraft”; the German verb impfen still means the same. An offshoot, naturally, makes one think of a child, and indeed imp soon came to mean “scion.” It is the next step, the narrowing or specification of its meaning from “small creature” to “son of the Devil” or “petty devil,” that could not have been predicted. Nor did brats fare too well. The origin of brat is obscure. The most common hypothesis derives this contemptuous word for “child” from Irish brat “ragged garment.” Secure parallels of such a semantic leap are few, and in ragamuffin rag- has probably nothing to do with ragtag and bobtail. Still another young male of questionable integrity is a shaver. A barber who can shave close to the skin is a sharper or a shaver. Today a (young) shaver is not necessarily an artful dodger, but the word, though it means “young lad,” is informal and has unquestionable ironic connotations.
Some stories have a happy end. Thus, one more Russian word for “child” is rebyonok (stress on the second syllable). The root reb– is the same as in orphan (with metathesis: in one word r precedes, in the other follows the vowel). Latin orbus means “bereft,” and Russian rab means “slave” (bereave is not related to them). Compare Engl. boy and busboy: “boy” and “servant” often form an alliance. It is reassuring that rebyonok is neither an orphan nor a slave. The topic is inexhaustible. Think of stripling (a small strip, as it were) and a slip of a girl (boy). Children vary, and so do the metaphors for their names.