Oxford University Press's
Academic Insights for the Thinking World

Ice: a forlorn hope

Why a forlorn hope? Because all the Germanic-speaking people had the same word for “ice,” and yet we don’t know where it came from. Some of our readers may be aware of the statement that the Eskimo (Aleut) language has fifty (just fifty!) words for “snow.” This statement aroused a long controversy, which is none of our business, but, in principle, it should not surprise anyone that the better we are acquainted with an object, the more nuances we can recognize in it and the more words will or may be coined to describe its features: compare the endless number of color words in a painter’s vocabulary. Icelandic, a language spoken by the people who have dealt with cold since the day they settled in their country (nor were they ignorant of it in Norway!), distinguishes between “ice in general,” “hollow ice,” “ice on a skating rink,” and “ice as frozen ground.” We tend to believe that the ancient Germanic word ice, from īs, was a generic term denoting “frozen water.” This belief may be true but may be an illusion.

Not exactly an icicle. (Image by Alto Crew via Unsplash)

The speakers of Medieval Icelandic and German did not ponder the etymology of īs, but they heard that īs sounded very much like īsarn, their word for “iron.” (Though the Old English for “iron” was īren, the protoform mut have had s in the root.) For a substance to be called “ice,” it had to be cold, hard, and under some circumstances glittering or gleaming. Old Icelandic poetry testifies to an almost complete merger of íss “ice” and ísarn “iron, metal” in some contexts. In that poetry, another word for “ice” (svell “hollow ice”) functioned as a synonym for “a gleaming sword.” Curiously, in Old English, the word ice especially often occurred in connection with melting, and in the poetry of that period, swords tended to melt (this once happened to Beowulf’s weapon)! Scyld Scefing, the great ancestor, extolled in Beowulf, died and was given a splendid sea burial. His ship, ready for departure, is called īsig. Perhaps “all aglitter”? “Icy” makes no sense. Despite all that, ice cannot be shown to be related to iron. (The etymological connection between ice and iron has been tentatively suggested in several sources, including The Century Dictionary, usually a reliable reference work.) Words that are pronounced in a similar way but have different meanings are called paronyms. Such were the Old Germanic names for “ice” and “iron.”

Ice and steal may gleam but the words for them are not related. (Image by One lucky guy via Flickr)

Those who will take the trouble to look up ice in etymological dictionaries will read that this word occurred in all the Germanic languages except Gothic. From fourth-century Gothic we have part of the New Testament. Icebergs and frozen rivers cannot be expected to grace the Biblical landscape. Yet in the Old Testament, ice does occur more than forty times (especially often as part of the compound ice-storm). Most of them are concentrated in Exodus, but we also read in Job: “My brethren have dealt deceitfully as a brook, and as the stream of brooks they pass away, which are blackish by reason of the ice, and wherein the snow is hid…” VI: 15-16. (Snow and ice… Why blackish?) Also, in Job 38: 29: “…out of whose womb came the ice.” Nor are Psalms free from ice (for instance: “He casteth forth his ice like morsels”—147: 1; like morsels? All the passages are from the Revised Version), but only once do we come across ice in the New Testament (Revelations). So even if all four gospels had been extant in Gothic, we would still have not found the word in that text. (I’ll stay away from the ludicrous idea that Jesus walked on the ice.) Yet a Gothic rune has come down to us. It has the strange name iiz. No one knows the value of the last letter. Yet the rune certainly means “ice.” Thus, īs was indeed the Common Germanic word for “ice.”

When in an ancient language a long vowel occurs before s, there is a suspicion that once the word began with a short vowel followed by n or m (the nasal consonant was allegedly lost, and the vowel acquired length by compensation, as, for instance happened in English five, which once had in or im in the root: compere German fünf). Germanic īs may have at one time begun with in-, but this idea was refuted almost as soon as it was offered. Yet one can see references to it in some good modern dictionaries, even though presented hesitatingly. An often-cited cognate of ice is Russian inei “rime.” However, the origin of inei has not been discovered. In this blog, I keep repeating that one word of unknown etymology cannot be called upon to elucidate another equally or even more obscure word (obscurum per obscurius).

Ice has a few lookalikes in Celtic (in Gaelic and Irish; for instance, Middle Irish aig). Some words for “ice” that occur in the Iranian languages (isu-, aēxa, yex, and so forth; read x as kh) also have a similar shape: a long vowel and a guttural sound after it. The similarity between all of them and Germanic īs (or its putative Indo-European form eis) can hardly be called striking. Basque izos “frost” begins with iz– and appears to be a good match for Germanic īs, but why should the speakers of Germanic have borrowed a Basque name for “ice”? This etymology has been proposed within the framework of a hypothesis that looks upon numerous words of questionable origin as loans from Basque. By contrast, in Latin, we find only glaciēs “ice,” familiar from English glacier and German Gletscher. This is a typical Alpine word, like English avalanche and possibly German Lawine (the same meaning). Such words made it into our modern languages from Alpine dialects, and their origin can seldom, if ever, be recovered.

What grows with its head down? (Image by Kiwihug via Unsplash)

Several fanciful etymologies of ice exist. They need not concern us here: unproductive guesswork. More important is the fact that Old Icelandic jaki “ice floe” (the root of jökull) is supposed to be related to the Celtic words mentioned above (for instance, Middle Irish aig “ice”). If so, jaki and ice emerge as closely related, almost as doublets. Is icicle a tautological compound “ice-ice”? (See the posts for 21 June 2006 and 26 July 2006, in which such words are discussed.) This conclusion does not sound fully convincing. Perhaps Germanic did have at least two words for “ice.”

As usual, when rambling becomes confusing, it may be worthwhile to look back and draw some conclusions, however tentative. In dealing with the etymology of ice, we have encountered several conjectures. Perhaps ice is not an Indo-European noun but a so-called migratory word (German Wanderwort) for “frost” or some variety of ice (solid ice, hollow ice, packed ice, crushed ice, iceberg, glacier, or whatever), later generalized. Or perhaps it was coined far away from the historical habitat of the Scandinavians and Germans and reached them along with many other words later used to describe their native landscape. Those who adopted īs seem to have some doubts about how to use it: the noun was masculine in Scandinavian but neuter in West Germanic! Or ice might be a word that had a transparent origin, but we have no way of guessing it.

Feature image by Aleksi Aaltonen via Flickr

Recent Comments

There are currently no comments.