An Etymological Dog Kennel, or Dog Eat Dog: Icelandic setja upp við dogg, Engl. to lie doggo, Engl. dog, and Engl. it’s raining cats and dogs

There is a rare slangy British English phrase to lie doggo “to lie hid”. The earliest known example is dated in the OED to 1882. Doggo looks like dog + o (with -o, as in weirdo, typo, and so forth), but a formation consisting of an animal name followed by the suffix -o would have no analogs. Some light on the origin of lie doggo may fall from the Modern Icelandic idiom sitja upp við dogg ‘to sit or half-lie, supporting oneself with elbows’. Doggur, known from texts since the eighteenth century, occurs with several other verbs. Also, sitja eins og doggur ‘sit motionless, look distraught’ and vera eins og doggur ‘to be motionless’ exist. Doggur has nothing to do with dogs, because the Scandinavian word for “dog” is hund-. The origin of the English noun dog is obscure, but, contrary to the almost universal opinion, the word is not totally isolated. In some German dialects, the diminutive forms dodel, döggel, and the similar-sounding tiggel ~ teckel occur. Perhaps dog and its continental look-alikes were originally baby words. The same sound complexes as above sometimes mean ‘a cylindrical object’ (such are Icelandic doggur and Middle High German tocke). Two of the basic meanings of those words were probably ‘round stick; doll’. Although the evidence is late, we can risk suggesting that lie doggo also contains the name of some device that was current not too long ago in the European itinerant handymen’s lingua franca. The overall image looks nearly the same as in the phrase dead as a doornail. In English, folk etymology connected doggo with the animal name and misled even professional lexicographers and etymologists. Finally, of some relevance is the English idiom it rains cats and dogs, whose forgotten earliest form was it rains cats and dogs and pitchforks with their points downwards. Apparently, the original idea was that a downpour of sharp objects fell to the ground. Shaping the Rings of the Scandinavian Fellowship. Festschrift in Honour of Ērika Sausverde. Edited by Ieva Steponavičiūtė Aleksiejūnienė and Loreta Vaicekauskienė. (Scandinavistica Vilnensis 14). Vilnius University Press, 2019. https://doi.org/10.15388/ScandinavisticaVilnensis.2019.10 Copyright © 2019 Authors. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Anatoly Liberman 202 1. Etymological Discovery and Luck Nineteenth-century historical linguists had every right to celebrate its victories. Thanks to Rasmus Rask, Jacob Grimm, and their followers, it became possible to lay bare the ties, previously not only hidden, but even unthinkable. At the crest of the Neogrammarian euphoria, the phrase the science of etymology gained ground, and a science it was, though it shared the weaknesses of every field of knowledge: the foundation looked and often proved to be solid, even though the buildings erected on it developed multiple cracks. Etymological algebra, which works well in principle, left many questions unanswered: it went only “so far”. Hundreds of entries in our best etymological dictionaries still contain no answers. As a reaction to this state of affairs, at the beginning of the twentieth century, a romantic school sprang up (Leo Spitzer and other Romance scholars) that emphasized the role of inspiration in etymology. Great discoveries are always inspired, but not every discovery is great, and in reconstructing the past, the line between inspiration and fantasizing is sometimes hard to draw. Looking back at my long experience, I can single out several non-trivial situations that result in finding a feasible solution. 1) A word has been given up as hopeless, but a convincing guess exists, hidden in an obscure place. This is how I came to know the origin of the English word slang. It was explained more than a hundred years ago, but the short article on the subject appeared in a local journal, not devoted to language problems. I ran into it twice: first, because my team and I screened all available resources for my etymological database (Liberman, 2010), and at the same time through a footnote in a relatively recent but also “well-hidden” work. A similar rediscovery happened in my dealing with Engl. dwarf and yet. Dwarf is an especially sad case. Friedrich Kluge came close to understanding the distant past of this word, but later gave up his suggestion, and it fell into oblivion. I rediscovered his train of thought, after which connecting the dots (from dwarf < *dwezg to dizzy) was not too hard. The etymology of yet came to light in a long article by an outstanding German specialist in Slavic (in a section of that article), and English language historians had little chance of noting it. 2) The etymology is almost obvious, but, to put two and two together, special circumstances are required. A parade example of this situation is the origin of Engl. pimp. German scholars explained the derivation of Pimpf ‘a little boy’ long ago (the word is almost transparent), but they did not know Engl. pimp. Even if they had been familiar with English books about pimps and prostitutes, they would have discovered only a An Etymological Dog Kennel, or Dog Eat Dog 203 curious analog of Pimpf and left it at that. Unfortunately, English philologists, in turn, had no notion of Pimpf and overlooked the only cognate that would have explained how pimp was coined (see the etymology of slang, dwarf, yet, and pimp in Liberman, 2008). Not only lay people but also linguists have trouble overcoming language barriers. I can cite one more example of the same type as pimp – Pimpf. English etymologists are not certain whether fog ‘mist’ and fog ‘second crop of grass; aftermath’ are connected. If they were aware of the Russian words par ‘steam’ and par ‘a field left unsown for one year’, they might have felt more secure. In Slavic, the connection between par1 and par2 is obvious (from a historical point of view, rather than in today’s linguistic intuition). An etymologist faces a haystack, and the needle may be hidden in Polish, Irish, or French, or, worse still, in some dialect. One cannot explore, let alone know, them all. Buck (1949) and its likes provide some help, but their range is relatively narrow. And this brings me to my last point. 3) Serendipity. The English word galoot means ‘raw soldier or mariner; uncouth man’. The few guesses about its origin are so unconvincing that they do not deserve reproducing. While hammering away at my bibliography (see above), I examined tens of thousands of pages. Once I ran into Öhman (1940). The title of the article looked rather unpromising, but I had such a high opinion of Öhman’s work that I decided to read that contribution, and on p. 149 saw Italian galeotto ~ galeoto ~ galioto ‘sailor’, corresponding to Old French galiot ‘pirate’. The word provided a perfect etymon of galoot (apparently, it was part of the sailors’ lingua franca). I published this discovery in my weekly blog The Oxford Etymologist, and it has gained some recognition. On the same website, I posted my comparison of two slang idioms: Engl. it gets on my wick ‘it irritates me’ and German es geht mir auf den Wecker (the same meaning). If I, by mere chance, had not known both, both would have remained impenetrable. There is some difference between the situations described at (1) and (2) – slang, dwarf, yet; pimp – and the two last ones: galoot and wick. No acceptable etymologies of galoot and wick have ever been proposed, so that they did not have to be rediscovered or excavated. I stumbled into viable solutions by chance. In scholarly work, chance is inseparable from a goal-oriented effort, but it still remains chance; hence my respectful attitude toward serendipity, and it is to serendipity that I’ll devote the rest of this paper, dedicated to Ērika Sausverde, with her talent for finding great things “accidentally” and putting them to use. Anatoly Liberman 204 2. The English Idiom to lie doggo I cannot remember when or where I came across the rare English idiom to lie doggo ‘to lie quiet, remain hid; without moving or making a sound; without doing anything that could draw attention’. That phrase did not show up in the first edition of the Oxford English Dictionary (OED); it is not entered at doggo or at lie. The idiom must have been equally foreign to American English in 1890–1910, for it is absent from the extremely detailed Century Dictionary. After the completion of the OED, a supplement was published, and in it, lie doggo is represented by three citations, dated to 1893, 1916, and 1924. The voluminous post-World War II supplement added nothing new, except a 1955 example. But the OED online has two pre-1893 citations, the earliest being dated to 1882. It is instructive to look at the evidence of doggo in some of our authoritative dictionaries. Although, most likely, no one knows anything definite about the word’s derivation, the attempts to disguise ignorance exist. The third edition of The Shorter Oxford Dictionary (1933) did not include the troublesome word, but that edition was revised many times. In 1942 and 1947 (=1952), the word appeared without any etymology (one finds only [?]). I have not seen the fourth and the fifth edition. Yet the sixth says that doggo occurs chiefly in lie doggo. The word chiefly is confusing, because the alternatives are not given. From the OED online we learn that play doggo also exists. In the first (1966) edition of the American Random House Dictionary, doggo is defined as ‘in concealment, out of sight’, and lie doggo, presumably from dog + o is called informal British slang. In the second (1983/1993) edition, the following example is given: “Lie doggo until the excitement blows over”. I suspect that the sentence was constructed by the editorial staff. With regard to the origin, all sources that risk an opinion say: “dog + (the suffix) -o”. The Oxford Dictionary of English Etymology (the 1985 reprint) adds a short comment: “[...] app[arently] f[rom] dog +o, with ref[erence] to the light sleeping of dogs and the difficulty of telling when their eyes are shut whether they are asleep”. Webster’s New International (WNI1) (1909) could not know anything about doggo, but the second edition (1949) contains a novelty. Definition: “Quietly out of sight, esp. in concealment of a trained dog”; this is followed by: “From the pretended sleep of a trained dog”. Where did the information about trained dogs come from? The usually careful Weekley (1921) wonders: “?Like a cunning dog”. The third edition of WNI (1981) admits that the origin is unknown, “probably from dog + o” and defines the word as slang: An Etymological Dog Kennel, or Dog Eat Dog 205 “Quietly out of sight, esp. in concealment – used chiefly in the phrase to lie doggo”. In the sixth edition of The Shorter Oxford, the formulation is the same as in the Random House Dictionary: “Used chiefly in lie doggo”. The suffix or ending o indeed exists. It occurs without any obvious function in cheerio; for adding a familiar ring to some forms of address (my boyo); in weirdo ‘an odd “weird” person’, a synonym for weirdie; for making words sound “folksy”, as in the name of Drano, that is, “draino” (a liquid for opening clogged sinks and toilets), and the like (compare combo and condo). There can be no doubt that Godot in Waiting for Godot is God + o. It has also been known for a long time that the element o enjoys special popularity in Australian English, where it is ubiquitous; the word kiddo ‘kid, child’ is a common occurrence in fiction. Quite recently, the Internet began using doggo as a meme for dog, and the word, as expected, “went viral”. This was the rebirth of doggo. The meme has nothing to do with the old idiom, and the ending (or suffix) -o does not prove that this doggo originated in Australia. As regards the slang phrase lie doggo, a certain detail may be considered. One of the earliest (1893) citations in the OED is from Kipling. Today, Kipling’s relatively short-lived sky-high fame is almost inconceivable: every word he wrote was noticed, remembered (it would pass into proverb!), and repeated throughout the English-speaking world. Perhaps lie doggo became more widely known from him. Kipling, who, like Dickens and Thackeray before him, enjoyed parading his knowledge of street language and slang, may have heard it from his military friends or picked it up from any part of the empire. If so, to lie doggo was not just slang, but possibly military slang, with the original sense ‘to lie quietly in ambush’. To be sure, the source may also have been hunters’ and sportsmen’s usage. Quite naturally, English speakers associated doggo with the animal name dog. However, dogs do not lie in ambush (they are usually urged to attack their prey). Nor does the idea that it is hard to know whether a dog is sleeping or awake sound too appealing. The questions of origins would have remained unassailable if there were not such a thing as serendipity. 3. Modern Icelandic sitja upp við dogg I came across the Icelandic phrase by chance, while reading a newspaper or a short story. It was new to me, but it is common enough, because all good dictionaries feature it. I immediately thought of Engl. lie doggo, Anatoly Liberman 206 which had bothered me for years, for its own sake and in connection with my attempts to understand the origin of the word dog. All the glosses and explanations below have been borrowed from ÁBM. Sitja upp við dogg (dogg is the accusative of doggur) means ‘to sit or half-lie, supporting oneself with elbows’. Unlike Engl. doggo, which, if one disregards the Internet meme, is a “bound word” and has not been recorded outside the idiom lie (play) doggo, Icel. doggur, known since the eighteenth century, has an independent existence and designates a vertical cylindrical object (let us not forget the definition). In addition to sitja upp, one can say liggja ‘to lie’, hallast ‘to hold oneself ’, and rísa ‘to rise’, with dogg after them. Two close variants include sitja eins og doggur ‘sit motionless; look distraught’ and vera eins og doggur ‘to be motionless, devoid of emotions’. The question is whether the idioms in Icelandic and English are related. Whatever the answer may be, one thing is clear: doggur has nothing to do with the animal name dog, because that animal is called hundur in Icelandic and Faroese, and hund in all the continental Scandinavian languages. A look at other Modern Icelandic words beginning with dogg reveals the following picture (the parentheses are again from ÁBM): doggast (late occurrence) ‘to do something mechanically, without giving thought to the matter, (? from doggur, as above) – [so not quite the same as “to pursue the matter doggedly”]; doggslegur ‘weak, feeble; sad; dull’ (no time of the earliest occurrence is given; hardly from *dokslegur: dok ‘doubt; hesitation; obstacle, impediment; drowsiness); dogginn (late) ‘persistent’, but also ‘quiet; depressed, down in the mouth’ (a phonetic variant of dokkin; dok ‘doubt, etc.’ – see it above; in some of its senses the word is geographically limited). The nouns doggur ‘dog’ and ‘young shark’ are recent borrowings from English. The conclusion offered in the entry runs as follows: “No cognates in Germanic, but cf. Norwegian dogg ‘boathook; gaff ’. The initial meaning may have been *‘pole, round stick; doll, puppet’”. One can summarize the evidence presented above so: Icel. doggur is late, and its origin is unknown; its only probable cognate is Norw. dogg, which, like doggur, designates an implement. However, the relevant objects differ: a vertical cylindrical object versus a boathook. Not improbably, gg in the root alternates with kk, but this conjecture cannot be substantiated. With this conclusion in view, it may be useful to examine briefly the history of Engl. dog. The most common statement in dictionaries is short: “Origin unknown”. However, the literature on dog is huge, the conjectures are many, An Etymological Dog Kennel, or Dog Eat Dog 207 and the findings in some more recent articles are significant. A few misleading look-alikes, such as Engl. dodge, Old Icel. dugga ‘a headstrong intractable person’, and the Old English personal name Dycga should probably be ignored. Speakers of Old English called the dog hund. In our earliest texts, only the form docgena, the genitive plural of the unattested word *docga appeared, and only a single time in the Boulogne Prudentius Glosses. It was applied contemptuously to the pagan henchman of the torturer Dacianus by their victim Vincentius and rendered Latin canum. The canum ~ docgena pair makes it certain that docgena did indeed mean ‘of (the) dogs’, and we note with surprise that the word’s pejorative sense turned up long before the regular, neutral one. It had either gained some currency as vulgar slang by the middle of the eleventh century and was avoided in writing or the glossator used an animal name current in his dialect, but unknown elsewhere (the glosses in question do contain rather many words not attested elsewhere). As regards word formation, *docga, with cg standing for gg, would have had the same structure as the recorded oldest forms of frog, stag, (ear)wig , and less certainly, pig. Those were hypocoristic words of the same type, though with a different suffix, as in Modern Engl. doggie and froggie. I will pass by the rather numerous but futile attempts to find Greek and Latin cognates of dog. By contrast, the related forms in Germanic are worthy of attention. Many modern languages have the word dog, and, whenever it appears, it is said to be a borrowing of Engl. dog. This pronouncement should not be accepted dogmatically. (It is hard to believe, but some brave amateurs did derive dog from dogma!) Today, few people consult the works of the once renowned German philologist August Pott (1802–1887). One of his many publications is an essay on dogs (Pott, 1863). He cited an astounding number of words for “dog”, among which we find dodel (apparently recorded in a German dialect of Alsace; the reference is unclear) and döggel from Schleswig. It is often impossible to determine the age of the otherwise unknown local and isolated words. They may be old or late, native or borrowed. Engl. dog ousted hound, that is, hund, in Middle English, when it became a generic animal name, with hound retaining the sense ‘hunting dog, dog kept for the chase’. Döggel may be a diminutive form taken over from English, but there is no certainty. In any case, dodel looks native. Even more instructive than Pott’s is the work of Werner Flechsig (1966), who investigated the name for “bitch” in Ostfalia (Low Saxony). Anatoly Liberman 208 In his discussion of the word Tache and its eighteenth-century synonym Tiggel, he suggested that they might be cognate with dog. Tiggel resembles döggel: both are diminutive forms, like dodel, and belong with the Old English animal names ending in cga (see above). Pott mentioned Teckel. Most likely, dog arose as a baby word. Not to be missed are two studi es: by Ulrike Roider (1982) and Thomas Markey (1983). Roider listed German dialectal dogke ‘a foolish woman; doll’, related to German Docke ‘doll’. Old dolls were often short pieces of wood dressed like manikins. One of the senses of Engl. dock is ‘the solid fleshy part of a horse’s tail; crupper, rump’ (hence the verb dock ‘to remove the end of the tail; cut short’). Roider suggested that Engl. dog is allied to Docke and that it got its name from the practice of docking dogs’ tails. There must have been a less tortuous path from “doll” to “dog”. Children regularly use the same monosyllable for a variety of objects, including toys and pets: the object’s form and function do not matter. Presumably, dog and its phonetic variants had limited currency in English and Low German; yet dog eventually edged out hound. Markey (1983) cited similar Low German words in his discussion of dog. They mean “girl; doll; clump, straw bundle”. Like Roider (1982), whose work he had naturally not read, he referred to breeds of dogs with artificially abbreviated tails. But I think he, too, erred and that he went astray in following Eric Hamp’s etymology of pig and reconstructing the basic meaning of all those words as “small, young”. Hamp connected Engl. pig and Danish pige ‘girl’. Both words are of unclear origin, and nothing harms a sought-for etymology more than attempts to explain an obscure word by referring to another obscure one. Incidentally, pig is also an upstart; it superseded swine. In pige, g goes back to k (late Old Icel. pika; the connection with Finnish piika has not been clarified to everybody’s satisfaction). To reiterate my tentative conclusion: dog is a baby word, which in the beginning had a wide range of application, from ‘doll; puppet’ to ‘pet; dog’. Pig might be a close neighbor (d g ~ p g and any vowel in the middle). Here are a few more words of the type discussed above. In Modern German, Docke means ‘a ball of thread; bundle’, ‘doll’ is dialectal (the Standard word for “doll” is Puppe). Old High German tocka ~ tocha has been recorded with the sense ‘puppet’, but Middle High German tocke meant ‘puppet; girl; cylindrical object; windlass; bunch’. Especially surprising is ‘cylindrical object’, because, as we have seen, one of the senses of Icel. doggur is also ‘vertical cylindrical object’. In the middle of the fifteenth century, Engl. dog began to be used for various “mechanical An Etymological Dog Kennel, or Dog Eat Dog 209 devices, usually [!] having or consisting of a tooth or claw, used for gripping or holding” (the original formulation in the OED). Despite the word usually, all the devices listed in the entry are indeed for gripping, drawing, and holding (hooks, bars, screws, nippers, and so forth). Some of the implements are called ‘cat’, and we’ll return to this fact below. The metaphor behind the name (“a thing with teeth”) is obvious. The same idea must have occurred to people everywhere (cf. Russian sobachka ‘trigger’, literally, ‘little dog’ – a borrowing of French chien or an independent formation?). In the Scandinavian languages, outside Modern Icelandic, one observes Old Icel. dokka ‘windlass’ (a late borrowing from Low German), ‘doll’; Norwegian dokke ~ dukke ‘doll’; Swedish docka (< dokka), and Danish dukke ‘short column; doll; puppet’. Norwegian dogg ‘boathook, gaff ’ has already been mentioned. ÁBM suggests the initial meaning of Icel. doggur to have been *‘pole, round stick; doll, puppet’. 4. A Preliminary Approach to the Origin of sitja up við dogg and Engl. lie doggo ÁBM’s set of heterogeneous reconstructed meanings (‘pole, round stick; doll, puppet’) causes surprise, but a broader look at the evidence presented above failed to yield better results. The Middle High German glosses of Docke (‘puppet; straw bundle’ and ‘staircase support’) point to a similar union of soft and hard objects. It seems that any round (“cylindrical”) thing, long and sturdy structure (a staircase support, a windlass), tall mannequin (doll), and even some pet animal (puppy) could be called dog or dock; the frequent use of diminutive suffixes points to baby language. The confusion of g and k makes our wanderings among such words especially hard, because this confusion is predictable only in German. (A brief digression: not only Engl. dock ‘stump, etc.’, with its protomeaning *‘something round’ is obscure; for dock ‘herb’ and dock ‘a creek in which a ship rests’ dictionaries can only cite cognates; dock ‘an enclosure in a criminal court’ is impenetrable.) There were two ways from the ancient symbiosis. One led to the establishment of animal names; it was taken by English and some German dialects. The other way led to the naming of devices, predominantly, but not necessarily, round. In English, the records of dog ‘device, implement’ do not antedate the late Middle period. Perhaps earlier occurrences are unknown to us, but, more probably, the words listed in the OED (see Anatoly Liberman 210 them above) reached England from the continent as part of the itinerant artisans’ lingua franca. I have discussed two such words in the past: adz(e) (in the 2008 dictionary) and ajar in Liberman (2014). The records of the Icelandic words are so late that all hypotheses are doomed to remain guesswork. Even in Icelandic, the merger of dog and dok is possible, especially in such words with voiceless sound groups as doggslegur ‘weak’. We occasionally note an unexpected combination of senses: cf. Icel. dogginn ‘persistent’ and ‘depressed’, ? from ‘staying in the same state’. Building semantic bridges is easy; that is why they collapse with such regularity. One gets the impression that some dog ~ dok adjectives could be used loosely, with semantic borders shifting or blurred, while the related nouns functioned like synonyms of Engl. thingamabob. Whatever the etymology of such words, sitja up (liggja, hallast, rísa) up við dogg has nothing to do with the animal name dog. The English case is less clear. Perhaps some light comes from the simile dead as a doornail. It seems to have been coined in allusion to nails driven deep, to the tack, into wood and remaining immobile. Some such image, evoked by doggur ‘a long round vertical object’, may have been at play in Icelandic, regardless of whether doggur was native or, more likely, a borrowing (or an adaptation) of a Low German word. Judging by final o in doggo, Engl. doggo was hardly coined in England. The late emergence of the phrase in Kipling points in the same direction. The source might be, as presumably in Icelandic, some Low German dog form, and it too may have been part of North Sea Germanic handymen’s professional argot. The connection with the habits of sleeping dogs is, in my opinion, due to folk etymology. The association of objects made for gripping and pulling with dogs is common in many parts of Europe, but this fact should not undermine the suggestion that the name of the implement existed parallel to, rather than as a derivative from, the animal name. 5. The English idiom it rains cats and dogs As is well-known, the idiom means ‘it pours with rain’. Its origin keeps puzzling speakers, but most of the explanations offered in the past should be discarded as fanciful. The phrase has been traced to French cantaloupe ‘waterfall’ (then a product of either a naïve pronunciation or folk etymology); to Greek katà doxas (? ‘down in full force’); to the fact that torrential rains carried along with them the refuse of the streets, including many dead animals; to the Scandinavian myth in which a cat is An Etymological Dog Kennel, or Dog Eat Dog 211 said to be influenced by the coming storm (there is no such myth!); to the superstition that cats were transformed witches, while dogs were the hounds of Odin (Óðinn), the god of storms (however, Óðinn was not a thunder god or a weather god and had no such hounds); to cats and dogs pattering across a bare boarded floor, strongly resembling the sound of a heavy downpour of rain; to Italian tempo cattivo ‘bad weather’; to the word catodupe, meaning a cataract (“the great katadoupeo – the cataracts of the Nile, from katadoupeo – to fall with a heavy sound. It is raining cats and dogs – it is raining cataracts”). It will be seen that many authors were satisfied by accounting for cats, as though the second word of the idiom (dogs) did not matter. However, it was known rather early that the full (original) text of the idiom was it rains cats and dogs and little pitchforks, alternating with it rains cats and dogs and pitchforks with their points downwards. The note by N. E. Toke (1918) deserves being reproduced almost in full: The “New English Dictionary” [= OED], under the heading “cat” 17, quotes G. Harvey, “Pierce’s Super”, 8 (1592), “Instead of thunderboltes shooteth nothing but dogboltes or catboltes”. This seems nearer the mark, but it is impossible to judge without the context, and this I do not know. By the way, “dogbolts” and “catbolts” are terms still employed in provincial dialect to denote, respectively, the iron bars for securing a door or gate, and the bolts for fastening together pieces of timber (see “English Dialect Dictionary” [by Joseph Wright]). A variety of the very popular game of trap and ball was called provincially “cat and dog” – the “dog” being the club with which the players propelled the “cat”, i.e. the piece of wood which, as in the game of tip-cat, did duty for the ball. If a number of players were engaged in this game and they grew excited, it might easily be said that it “rained cats and dogs” on the playing field. Could the expression have arisen in this way? A “dog” also means a portion of a rainbow, and generally precedes or accompanies squall at sea. In this connection, the “English Dialect Dictionary” quotes “It’ll mebbe be fine i’ t’efternoon if t’ thunner keeps off, but there’s too many little dogs about” (West York[shire]). The connection of “dogs” with a downpour of rain is accounted for by this use of the word. Some humorists may have added “cats”, and the phrase, thus originated, may have caught the popular fancy. But this is merely a suggestion, and I should be glad of a less hypothetical explanation. Anatoly Liberman 212 The OED offers the closest approximation to the phrase dated to a. 1652. The quotation from Harvey is now dated to 1593 (at dog bolt). The reference to the rainbow is curious, but it does not account for cats in the idiom. Also, the evidence of pitchforks (with their points downward) cannot be shaken off. Apparently, the original idea was that a downpour of sharp objects fell to the ground. Later, catbolts and dogbolts lost their second elements, and the phrase became more elegant but senseless. I hope that the history of the idiom will reinforce the idea that in the enigmatic phrase lie doggo, the reference is to some inanimate object. Perhaps the phrase was coined by players in a popular game rather than by wandering handymen (note dog ‘club’; reminiscent of the supports and cylindrical objects, mentioned above). We may never know. The important thing in this case is to let the sleeping dogs lie and not bother etymologists.


Etymological Discovery and Luck
Nineteenth-century historical linguists had every right to celebrate its victories. Thanks to Rasmus Rask, Jacob Grimm, and their followers, it became possible to lay bare the ties, previously not only hidden, but even unthinkable. At the crest of the Neogrammarian euphoria, the phrase the science of etymology gained ground, and a science it was, though it shared the weaknesses of every field of knowledge: the foundation looked and often proved to be solid, even though the buildings erected on it developed multiple cracks. Etymological algebra, which works well in principle, left many questions unanswered: it went only "so far". Hundreds of entries in our best etymological dictionaries still contain no answers. As a reaction to this state of affairs, at the beginning of the twentieth century, a romantic school sprang up (Leo Spitzer and other Romance scholars) that emphasized the role of inspiration in etymology. Great discoveries are always inspired, but not every discovery is great, and in reconstructing the past, the line between inspiration and fantasizing is sometimes hard to draw.
Looking back at my long experience, I can single out several non-trivial situations that result in finding a feasible solution.
1) A word has been given up as hopeless, but a convincing guess exists, hidden in an obscure place. This is how I came to know the origin of the English word slang. It was explained more than a hundred years ago, but the short article on the subject appeared in a local journal, not devoted to language problems. I ran into it twice: first, because my team and I screened all available resources for my etymological database (Liberman, 2010), and at the same time through a footnote in a relatively recent but also "well-hidden" work. A similar rediscovery happened in my dealing with Engl. dwarf and yet. Dwarf is an especially sad case. Friedrich Kluge came close to understanding the distant past of this word, but later gave up his suggestion, and it fell into oblivion. I rediscovered his train of thought, after which connecting the dots (from dwarf < *dwezg to dizzy) was not too hard. The etymology of yet came to light in a long article by an outstanding German specialist in Slavic (in a section of that article), and English language historians had little chance of noting it.
2) The etymology is almost obvious, but, to put two and two together, special circumstances are required. A parade example of this situation is the origin of Engl. pimp. German scholars explained the derivation of Pimpf 'a little boy' long ago (the word is almost transparent), but they did not know Engl. pimp. Even if they had been familiar with English books about pimps and prostitutes, they would have discovered only a curious analog of Pimpf and left it at that. Unfortunately, English philologists, in turn, had no notion of Pimpf and overlooked the only cognate that would have explained how pimp was coined (see the etymology of slang, dwarf, yet, and pimp in Liberman, 2008).
Not only lay people but also linguists have trouble overcoming language barriers. I can cite one more example of the same type as pimp -Pimpf. English etymologists are not certain whether fog 'mist' and fog 'second crop of grass; aftermath' are connected. If they were aware of the Russian words par 'steam' and par 'a field left unsown for one year', they might have felt more secure. In Slavic, the connection between par 1 and par 2 is obvious (from a historical point of view, rather than in today's linguistic intuition). An etymologist faces a haystack, and the needle may be hidden in Polish, Irish, or French, or, worse still, in some dialect. One cannot explore, let alone know, them all. Buck (1949) and its likes provide some help, but their range is relatively narrow. And this brings me to my last point.
3) Serendipity. The English word galoot means 'raw soldier or mariner; uncouth man' . The few guesses about its origin are so unconvincing that they do not deserve reproducing. While hammering away at my bibliography (see above), I examined tens of thousands of pages. Once I ran into Öhman (1940). The title of the article looked rather unpromising, but I had such a high opinion of Öhman's work that I decided to read that contribution, and on p. 149 saw Italian galeotto ~ galeoto ~ galioto 'sailor', corresponding to Old French galiot 'pirate' . The word provided a perfect etymon of galoot (apparently, it was part of the sailors' lingua franca). I published this discovery in my weekly blog The Oxford Etymologist, and it has gained some recognition. On the same website, I posted my comparison of two slang idioms: Engl. it gets on my wick 'it irritates me' and German es geht mir auf den Wecker (the same meaning). If I, by mere chance, had not known both, both would have remained impenetrable.
There is some difference between the situations described at (1) and (2) -slang, dwarf, yet; pimp -and the two last ones: galoot and wick. No acceptable etymologies of galoot and wick have ever been proposed, so that they did not have to be rediscovered or excavated. I stumbled into viable solutions by chance. In scholarly work, chance is inseparable from a goal-oriented effort, but it still remains chance; hence my respectful attitude toward serendipity, and it is to serendipity that I'll devote the rest of this paper, dedicated to Ērika Sausverde, with her talent for finding great things "accidentally" and putting them to use.
2. The English Idiom to lie doggo I cannot remember when or where I came across the rare English idiom to lie doggo 'to lie quiet, remain hid; without moving or making a sound; without doing anything that could draw attention' . That phrase did not show up in the first edition of the Oxford English Dictionary (OED); it is not entered at doggo or at lie. The idiom must have been equally foreign to American English in 1890-1910, for it is absent from the extremely detailed Century Dictionary.
After the completion of the OED, a supplement was published, and in it, lie doggo is represented by three citations, dated to 1893, 1916, and 1924. The voluminous post-World War II supplement added nothing new, except a 1955 example. But the OED online has two pre-1893 citations, the earliest being dated to 1882. It is instructive to look at the evidence of doggo in some of our authoritative dictionaries. Although, most likely, no one knows anything definite about the word's derivation, the attempts to disguise ignorance exist. The third edition of The Shorter Oxford Dictionary (1933) did not include the troublesome word, but that edition was revised many times. In 1942In and 1947, the word appeared without any etymology (one finds only [?]). I have not seen the fourth and the fifth edition. Yet the sixth says that doggo occurs chiefly in lie doggo. The word chiefly is confusing, because the alternatives are not given. From the OED online we learn that play doggo also exists. In the first (1966) edition of the American Random House Dictionary, doggo is defined as 'in concealment, out of sight' , and lie doggo, presumably from dog + o is called informal British slang. In the second (1983/1993) edition, the following example is given: "Lie doggo until the excitement blows over". I suspect that the sentence was constructed by the editorial staff.
With regard to the origin, all sources that risk an opinion say: "dog + (the suffix) -o". The Oxford Dictionary of English Etymology (the 1985 reprint) adds a short comment: "[…] app[arently] f[rom] dog +o, with ref[erence] to the light sleeping of dogs and the difficulty of telling when their eyes are shut whether they are asleep". Webster's New International (WNI1) (1909) could not know anything about doggo, but the second edition (1949) contains a novelty. Definition: "Quietly out of sight, esp. in concealment of a trained dog"; this is followed by: "From the pretended sleep of a trained dog". Where did the information about trained dogs come from? The usually careful Weekley (1921) wonders: "?Like a cunning dog". The third edition of WNI (1981) admits that the origin is unknown, "probably from dog + o" and defines the word as slang: "Quietly out of sight, esp. in concealment -used chiefly in the phrase to lie doggo". In the sixth edition of The Shorter Oxford, the formulation is the same as in the Random House Dictionary: "Used chiefly in lie doggo". The suffix or ending o indeed exists. It occurs without any obvious function in cheerio; for adding a familiar ring to some forms of address (my boyo); in weirdo 'an odd "weird" person', a synonym for weirdie; for making words sound "folksy", as in the name of Drano, that is, "draino" (a liquid for opening clogged sinks and toilets), and the like (compare combo and condo). There can be no doubt that Godot in Waiting for Godot is God + o. It has also been known for a long time that the element o enjoys special popularity in Australian English, where it is ubiquitous; the word kiddo 'kid, child' is a common occurrence in fiction.
Quite recently, the Internet began using doggo as a meme for dog, and the word, as expected, "went viral". This was the rebirth of doggo. The meme has nothing to do with the old idiom, and the ending (or suffix) -o does not prove that this doggo originated in Australia. As regards the slang phrase lie doggo, a certain detail may be considered. One of the earliest (1893) citations in the OED is from Kipling. Today, Kipling's relatively short-lived sky-high fame is almost inconceivable: every word he wrote was noticed, remembered (it would pass into proverb!), and repeated throughout the English-speaking world. Perhaps lie doggo became more widely known from him. Kipling, who, like Dickens and Thackeray before him, enjoyed parading his knowledge of street language and slang, may have heard it from his military friends or picked it up from any part of the empire. If so, to lie doggo was not just slang, but possibly military slang, with the original sense 'to lie quietly in ambush' . To be sure, the source may also have been hunters' and sportsmen's usage.
Quite naturally, English speakers associated doggo with the animal name dog. However, dogs do not lie in ambush (they are usually urged to attack their prey). Nor does the idea that it is hard to know whether a dog is sleeping or awake sound too appealing. The questions of origins would have remained unassailable if there were not such a thing as serendipity.
3. Modern Icelandic sitja upp við dogg I came across the Icelandic phrase by chance, while reading a newspaper or a short story. It was new to me, but it is common enough, because all good dictionaries feature it. I immediately thought of Engl. lie doggo, which had bothered me for years, for its own sake and in connection with my attempts to understand the origin of the word dog. All the glosses and explanations below have been borrowed from ÁBM. Sitja upp við dogg (dogg is the accusative of doggur) means 'to sit or half-lie, supporting oneself with elbows' . Unlike Engl. doggo, which, if one disregards the Internet meme, is a "bound word" and has not been recorded outside the idiom lie (play) doggo, Icel. doggur, known since the eighteenth century, has an independent existence and designates a vertical cylindrical object (let us not forget the definition). In addition to sitja upp, one can say liggja 'to lie' , hallast 'to hold oneself ' , and rísa 'to rise' , with dogg after them. Two close variants include sitja eins og doggur 'sit motionless; look distraught' and vera eins og doggur 'to be motionless, devoid of emotions' .
The question is whether the idioms in Icelandic and English are related. Whatever the answer may be, one thing is clear: doggur has nothing to do with the animal name dog, because that animal is called hundur in Icelandic and Faroese, and hund in all the continental Scandinavian languages.
A look at other Modern Icelandic words beginning with dogg reveals the following picture (the parentheses are again from ÁBM): doggast (late occurrence) 'to do something mechanically, without giving thought to the matter, (? from doggur, as above) -[so not quite the same as "to pursue the matter doggedly"]; doggslegur 'weak, feeble; sad; dull' (no time of the earliest occurrence is given; hardly from *dokslegur: dok 'doubt; hesitation; obstacle, impediment; drowsiness); dogginn (late) 'persistent' , but also 'quiet; depressed, down in the mouth' (a phonetic variant of dokkin; dok 'doubt, etc. ' -see it above; in some of its senses the word is geographically limited). The nouns doggur 'dog' and 'young shark' are recent borrowings from English. The conclusion offered in the entry runs as follows: "No cognates in Germanic, but cf. Norwegian dogg 'boathook; gaff' . The initial meaning may have been *'pole, round stick; doll, puppet'".
One can summarize the evidence presented above so: Icel. doggur is late, and its origin is unknown; its only probable cognate is Norw. dogg, which, like doggur, designates an implement. However, the relevant objects differ: a vertical cylindrical object versus a boathook. Not improbably, gg in the root alternates with kk, but this conjecture cannot be substantiated. With this conclusion in view, it may be useful to examine briefly the history of Engl. dog.
The most common statement in dictionaries is short: "Origin unknown". However, the literature on dog is huge, the conjectures are many, and the findings in some more recent articles are significant. A few misleading look-alikes, such as Engl. dodge, Old Icel. dugga 'a headstrong intractable person', and the Old English personal name Dycga should probably be ignored.
Speakers of Old English called the dog hund. In our earliest texts, only the form docgena, the genitive plural of the unattested word *docga appeared, and only a single time in the Boulogne Prudentius Glosses. It was applied contemptuously to the pagan henchman of the torturer Dacianus by their victim Vincentius and rendered Latin canum. The canum ~ docgena pair makes it certain that docgena did indeed mean 'of (the) dogs', and we note with surprise that the word's pejorative sense turned up long before the regular, neutral one. It had either gained some currency as vulgar slang by the middle of the eleventh century and was avoided in writing or the glossator used an animal name current in his dialect, but unknown elsewhere (the glosses in question do contain rather many words not attested elsewhere). As regards word formation, *docga, with cg standing for gg, would have had the same structure as the recorded oldest forms of frog, stag, (ear)wig, and less certainly, pig. Those were hypocoristic words of the same type, though with a different suffix, as in Modern Engl. doggie and froggie.
I will pass by the rather numerous but futile attempts to find Greek and Latin cognates of dog. By contrast, the related forms in Germanic are worthy of attention. Many modern languages have the word dog, and, whenever it appears, it is said to be a borrowing of Engl. dog. This pronouncement should not be accepted dogmatically. (It is hard to believe, but some brave amateurs did derive dog from dogma!) Today, few people consult the works of the once renowned German philologist August Pott (1802-1887). One of his many publications is an essay on dogs (Pott, 1863). He cited an astounding number of words for "dog", among which we find dodel (apparently recorded in a German dialect of Alsace; the reference is unclear) and döggel from Schleswig.
It is often impossible to determine the age of the otherwise unknown local and isolated words. They may be old or late, native or borrowed. Engl. dog ousted hound, that is, hund, in Middle English, when it became a generic animal name, with hound retaining the sense 'hunting dog, dog kept for the chase' . Döggel may be a diminutive form taken over from English, but there is no certainty. In any case, dodel looks native. Even more instructive than Pott's is the work of Werner Flechsig (1966), who investigated the name for "bitch" in Ostfalia (Low Saxony).
In his discussion of the word Tache and its eighteenth-century synonym Tiggel, he suggested that they might be cognate with dog. Tiggel resembles döggel: both are diminutive forms, like dodel, and belong with the Old English animal names ending in cga (see above). Pott mentioned Teckel.
Most likely, dog arose as a baby word. Not to be missed are two studi es: by Ulrike Roider (1982) and Thomas Markey (1983). Roider listed German dialectal dogke 'a foolish woman; doll' , related to German Docke 'doll' . Old dolls were often short pieces of wood dressed like manikins. One of the senses of Engl. dock is 'the solid fleshy part of a horse's tail; crupper, rump' (hence the verb dock 'to remove the end of the tail; cut short'). Roider suggested that Engl. dog is allied to Docke and that it got its name from the practice of docking dogs' tails. There must have been a less tortuous path from "doll" to "dog". Children regularly use the same monosyllable for a variety of objects, including toys and pets: the object's form and function do not matter. Presumably, dog and its phonetic variants had limited currency in English and Low German; yet dog eventually edged out hound. Markey (1983) cited similar Low German words in his discussion of dog. They mean "girl; doll; clump, straw bundle". Like Roider (1982), whose work he had naturally not read, he referred to breeds of dogs with artificially abbreviated tails. But I think he, too, erred and that he went astray in following Eric Hamp's etymology of pig and reconstructing the basic meaning of all those words as "small, young". Hamp connected Engl. pig and Danish pige 'girl' . Both words are of unclear origin, and nothing harms a sought-for etymology more than attempts to explain an obscure word by referring to another obscure one. Incidentally, pig is also an upstart; it superseded swine. In pige, g goes back to k (late Old Icel. pika; the connection with Finnish piika has not been clarified to everybody's satisfaction). To reiterate my tentative conclusion: dog is a baby word, which in the beginning had a wide range of application, from 'doll; puppet' to 'pet; dog' . Pig might be a close neighbor (d -g ~ p -g and any vowel in the middle).
Here are a few more words of the type discussed above. In Modern German, Docke means 'a ball of thread; bundle', 'doll' is dialectal (the Standard word for "doll" is Puppe). Old High German tocka ~ tocha has been recorded with the sense 'puppet', but Middle High German tocke meant 'puppet; girl; cylindrical object; windlass; bunch' . Especially surprising is 'cylindrical object' , because, as we have seen, one of the senses of Icel. doggur is also 'vertical cylindrical object' . In the middle of the fifteenth century, Engl. dog began to be used for various "mechanical devices, usually [!] having or consisting of a tooth or claw, used for gripping or holding" (the original formulation in the OED). Despite the word usually, all the devices listed in the entry are indeed for gripping, drawing, and holding (hooks, bars, screws, nippers, and so forth). Some of the implements are called 'cat' , and we'll return to this fact below. The metaphor behind the name ("a thing with teeth") is obvious. The same idea must have occurred to people everywhere (cf. Russian sobachka 'trigger', literally, 'little dog' -a borrowing of French chien or an independent formation?).

A Preliminary
Approach to the Origin of sitja up við dogg and Engl. lie doggo ÁBM's set of heterogeneous reconstructed meanings ('pole, round stick; doll, puppet') causes surprise, but a broader look at the evidence presented above failed to yield better results. The Middle High German glosses of Docke ('puppet; straw bundle' and 'staircase support') point to a similar union of soft and hard objects. It seems that any round ("cylindrical") thing, long and sturdy structure (a staircase support, a windlass), tall mannequin (doll), and even some pet animal (puppy) could be called dog or dock; the frequent use of diminutive suffixes points to baby language. The confusion of g and k makes our wanderings among such words especially hard, because this confusion is predictable only in German. (A brief digression: not only Engl. dock 'stump, etc.', with its protomeaning *'something round' is obscure; for dock 'herb' and dock 'a creek in which a ship rests' dictionaries can only cite cognates; dock 'an enclosure in a criminal court' is impenetrable.) There were two ways from the ancient symbiosis. One led to the establishment of animal names; it was taken by English and some German dialects. The other way led to the naming of devices, predominantly, but not necessarily, round. In English, the records of dog 'device, implement' do not antedate the late Middle period. Perhaps earlier occurrences are unknown to us, but, more probably, the words listed in the OED (see them above) reached England from the continent as part of the itinerant artisans' lingua franca. I have discussed two such words in the past: adz(e) (in the 2008 dictionary) and ajar in Liberman (2014).
The records of the Icelandic words are so late that all hypotheses are doomed to remain guesswork. Even in Icelandic, the merger of dog and dok is possible, especially in such words with voiceless sound groups as doggslegur 'weak' . We occasionally note an unexpected combination of senses: cf. Icel. dogginn 'persistent' and 'depressed' , ? from 'staying in the same state' . Building semantic bridges is easy; that is why they collapse with such regularity. One gets the impression that some dog ~ dok adjectives could be used loosely, with semantic borders shifting or blurred, while the related nouns functioned like synonyms of Engl. thingamabob.
Whatever the etymology of such words, sitja up (liggja, hallast, rísa) up við dogg has nothing to do with the animal name dog. The English case is less clear. Perhaps some light comes from the simile dead as a doornail. It seems to have been coined in allusion to nails driven deep, to the tack, into wood and remaining immobile. Some such image, evoked by doggur 'a long round vertical object' , may have been at play in Icelandic, regardless of whether doggur was native or, more likely, a borrowing (or an adaptation) of a Low German word. Judging by final o in doggo, Engl. doggo was hardly coined in England. The late emergence of the phrase in Kipling points in the same direction. The source might be, as presumably in Icelandic, some Low German dog form, and it too may have been part of North Sea Germanic handymen's professional argot. The connection with the habits of sleeping dogs is, in my opinion, due to folk etymology. The association of objects made for gripping and pulling with dogs is common in many parts of Europe, but this fact should not undermine the suggestion that the name of the implement existed parallel to, rather than as a derivative from, the animal name.

The English idiom it rains cats and dogs
As is well-known, the idiom means 'it pours with rain' . Its origin keeps puzzling speakers, but most of the explanations offered in the past should be discarded as fanciful. The phrase has been traced to French cantaloupe 'waterfall' (then a product of either a naïve pronunciation or folk etymology); to Greek katà doxas (? 'down in full force'); to the fact that torrential rains carried along with them the refuse of the streets, including many dead animals; to the Scandinavian myth in which a cat is said to be influenced by the coming storm (there is no such myth!); to the superstition that cats were transformed witches, while dogs were the hounds of Odin (Óðinn), the god of storms (however, Óðinn was not a thunder god or a weather god and had no such hounds); to cats and dogs pattering across a bare boarded floor, strongly resembling the sound of a heavy downpour of rain; to Italian tempo cattivo 'bad weather'; to the word catodupe, meaning a cataract ("the great katadoupeo -the cataracts of the Nile, from katadoupeo -to fall with a heavy sound. It is raining cats and dogs -it is raining cataracts"). It will be seen that many authors were satisfied by accounting for cats, as though the second word of the idiom (dogs) did not matter.
However, it was known rather early that the full (original) text of the idiom was it rains cats and dogs and little pitchforks, alternating with it rains cats and dogs and pitchforks with their points downwards. The note by N. E. Toke (1918) deserves being reproduced almost in full: The "New English Dictionary" [= OED], under the heading "cat" 17, quotes G. Harvey, "Pierce's Super", 8 (1592), "Instead of thunderboltes shooteth nothing but dogboltes or catboltes". This seems nearer the mark, but it is impossible to judge without the context, and this I do not know. By the way, "dogbolts" and "catbolts" are terms still employed in provincial dialect to denote, respectively, the iron bars for securing a door or gate, and the bolts for fastening together pieces of timber (see "English Dialect Dictionary" [by Joseph Wright]). A variety of the very popular game of trap and ball was called provincially "cat and dog" -the "dog" being the club with which the players propelled the "cat", i.e. the piece of wood which, as in the game of tip-cat, did duty for the ball. If a number of players were engaged in this game and they grew excited, it might easily be said that it "rained cats and dogs" on the playing field. Could the expression have arisen in this way? A "dog" also means a portion of a rainbow, and generally precedes or accompanies squall at sea. In this connection, the "English Dialect Dictionary" quotes "It'll mebbe be fine i' t'efternoon if t' thunner keeps off, but there's too many little dogs about" (West York[shire]). The connection of "dogs" with a downpour of rain is accounted for by this use of the word. Some humorists may have added "cats", and the phrase, thus originated, may have caught the popular fancy. But this is merely a suggestion, and I should be glad of a less hypothetical explanation.
The OED offers the closest approximation to the phrase dated to a. 1652. The quotation from Harvey is now dated to 1593 (at dog bolt). The reference to the rainbow is curious, but it does not account for cats in the idiom. Also, the evidence of pitchforks (with their points downward) cannot be shaken off. Apparently, the original idea was that a downpour of sharp objects fell to the ground. Later, catbolts and dogbolts lost their second elements, and the phrase became more elegant but senseless.
I hope that the history of the idiom will reinforce the idea that in the enigmatic phrase lie doggo, the reference is to some inanimate object. Perhaps the phrase was coined by players in a popular game rather than by wandering handymen (note dog 'club'; reminiscent of the supports and cylindrical objects, mentioned above). We may never know. The important thing in this case is to let the sleeping dogs lie and not bother etymologists.