This is the first in what I hope will be a (very sporadic) series on the roots of English. It may also be the last; I’ll have to see how it goes. It’s very difficult to write at all accurately on this subject without getting awfully technical, and I’m not sure that I can avoid getting mired in incomprehensible neepery. Caveat lector!
What do the words filth, length, and strength have in common?
Well, they’re all nouns. More specifically, they’re all abstract nouns of quality or state, and each is related to an adjective in much the way that loudness is related to the adjective loud or stupidity to stupid: foul, long, and strong.
And they all end in -th.
It’s almost as if someone added a suffix -th, with a meaning something like that of the suffixes -ness and -ity, to each of the adjectives and ran the resulting ‘words’ foulth, longth, and strongth through a wringer of some kind.
That’s not quite what happened, but it’s close. Late Proto-Indo-European (PIE) had a suffix *-éteh₂ that was added to adjectives to produce abstract nouns. In due course regular phonological change turned this into *-eþā, which then developed into the Proto-Germanic (PGmc) suffix *-iþō.
Okay, I’d probably better pause here for a brief digression. The asterisk denotes a form reconstructed by historical linguists; the odd-looking symbol *h₂ represents a reconstructed sound whose specific phonetic nature is still rather uncertain; among the likelier candidates are the voiceless and voiced pharyngeal fricatives and the voiceless uvular fricative. And you don’t have to worry about it, because it now leaves this story. The thorn (*þ) is used in Germanic historical linguistics for a phoneme reconstructed to have been a voiceless dental fricative, more or less the sound of th in thin and lithium (not the sound in this and lathe). The macrons indicate long vowels. Note that long here means long in duration (as compared with non-long vowels); it does not have the meaning that most of us in the U.S. learned in grade school, exemplified in pronunciations keys like this one in most U.S. dictionaries. Thus, both *a and *ā are reconstructed vowels something like ah (as in the doctor’s ‘Say aah’), but the second is longer in duration than the first.
Now back to PGmc, the last common ancestor of the Germanic languages, including, in particular, English. On the basis of cognates in the earliest attested Germanic languages historical linguists can reconstruct PGmc adjectives *fūlaz ‘foul’, *langaz ‘long’, and *strangaz ‘strong’. The first of these, for example, is inferred from Gothic fuls ‘fetid, foul’; Old Norse fúll ‘rotten, stinking’; Old English (OE) fúl ‘foul, dirty, rotten’, the ancestor of Present-Day English foul; Old Frisian fūl ‘stinking’, Middle Low German vūl ‘rotten, spoiled’, and Old High German fūl ‘foul, putrid’ and their declensions in their respective languages. Evidence for the others is similar, and I’ll spare you the gory details.
Adding the suffix *-iþō to these adjectives produced the PGmc nouns *fūliþō, *langiþō, and *strangiþō, meaning something like ‘filth, foulness’, ‘length’, and ‘strength’, respectively. The first of these, for instance, became fūlitha ‘rot, corruption’ in Old Saxon (whence Dutch vuilte ‘dirt’) and fūlida ‘filth, smut’ in Old High German. So what happened to it in the subsequent history of English to give us filth?
What happened is a sound change that is very common in the histories of the assorted Germanic languages, though the details vary from one language to another. This sound change is called I-mutation, I-umlaut, or front mutation. In this change a vowel i or ī, or a consonant j (representing the sound of y in yes) in a later syllable changed the vowel of an earlier syllable. For some reason speakers started to anticipate the tongue position needed to produce the i, ī, or j. Since these sounds are made with the front of the tongue blade (excluding the very tip) high in the mouth, right behind the upper front teeth, speakers were raising or moving the tongue forward too early, and this changed the quality of the preceding vowel.
In particular, I-umlaut occurred in the prehistory of Old English, sometime during its development from PGmc. The *i of the suffix *iþō caused I-mutation of the preceding vowel, so that the PGmc *ū of *fūliþō moved forward in the mouth and became Old English ý, a long vowel like that of French pur and German führen. (The acute accent is an editorial addition indicating that the vowel was long in Old English.) After that, but still before the earliest recorded Old English, the unstressed -i- of the suffix was lost and the final *-ō was reduced to *-u and then dropped, leaving OE fýlþ ‘filth, uncleanness, impurity’. This in due course became our filth as a result of regular developments from Old to Middle to Present-Day English.
Similarly, I-mutation moved the *a of *langiþō, and *strangiþō forward in the mouth to æ, the vowel of English cat, and then up to e, the vowel of French fée or the first vowel of German egal. The other changes on the way to Old English were the same as in fýlþ and resulted in OE lengþ and strengþ, which with little further change eventually became our length and strength. Variants of both with preserved final -u are also recorded, and we also sometimes find the earlier æ in the latter word: strængþ and strængþu. (Old English spelling was not standardized, and I’m ignoring variations that were merely orthographic.)
As an aside, I-mutation is why the plural of mouse is mice. The PGmc nominative singular and plural were *mūs and *mūsiz, respectively. The singular became OE mús ‘mouse’, which developed regularly into our mouse just as OE fúl developed into foul. The plural became OE mýs, which developed regularly into our mice. (You may be wondering why mice and filth have different vowel sounds if the vowels both developed regularly from OE ý. To oversimplify just a little, the difference is the result of a Middle English shortening of long vowels that were followed by two consonants.)
Okay, that’s probably more than enough technicalities; I can’t avoid them completely, but I’ll try to cut back a big. In particular, I’ll stop explaining the sound values of all of the symbols used in the Old English and Proto-Germanic; if anyone is especially curious, I’ll try to answer questions in the comments.
So where else does this fossilized suffix occur?
-
mirth, from OE myrȝþ ‘joy, pleasure’ in turn from PGmc *murgiþō; this was derived from PGmc *murguz, the source of OE myrȝe ‘pleasant, delightful, sweet’, which eventually became Present-Day English (PDE) merry.
-
truth, from OE tréowþ ‘truth; honesty, uprightness; loyalty, faithfulness’, in turn from PGmc *trewwiþō; the adjective was PGmc *trewwaz, the source of OE tréowe ‘loyal, trustworthy; honorable, honest; in accordance with fact or reality’, which eventually became PDE true.
-
height, from OE híehþo (later héahþo), in turn from PGmc *xauxiþō; the adjective was PGmc *xauxaz, the source of OE héah ‘high’, which eventually became PDE high. In Middle English some speakers started to replace the th sound at the end of the noun with a t sound, especially in the North of England. This change eventually became standard, though the noun was sometimes spelled heighth or highth down to the eighteenth century and is still sometimes pronounced that way today.
-
health, from OE hǽlþ ‘health’, in turn from PGmc *xailiþō; the adjective was PGmc *xailaz, the source of OE *hál ‘hale, whole, entire, uninjured’, which by different routes is the source of both PDE hale and PDE whole.
- Probably dearth, from Middle English derthe ‘a period of scarcity or high prices; a famine’. Although no such word survives in the extant Old English corpus, it is thought that there was an OE *díerþ(u) from PGmc *deuriþō, whose existence is assured by its reflexes (descendants) in Old Norse, Old Saxon, and Old High German. The corresponding adjective was PGmc *deurjaz ‘dear, precious, valuable, expensive’, whence OE díore and our dear.
-
theft, from OE þíefþ ‘theft; stolen goods’, in turn from PGmc *þeuβiþō. Here the noun-forming suffix *-iþō has been added not to an adjective, but to another noun, PGmc *þeuβaz ‘thief’, the source of OE þéof ‘thief’ and hence eventually of PDE thief.
Some apparent examples are misleading: breadth, for instance, only appears in the 1500s (as bredeth), replacing Middle English (ME) bræde ‘breadth, width’ from OE brǽdu. It was evidently formed from broad by analogy with words like length and strength. Width is even later: it doesn’t appear until the 1600s, when it started to replace widness ‘width, wideness’ (from OE wídnes ‘width’), by analogy with breadth, length, etc. Depth appears a bit earlier: Wyclif (14th century) has depthe, derived from ME dēp ‘deep’ (whence our deep) by analogy with length. These three nouns are semantically related to the adjectives broad, wide, and long in the same way that filth, length, and strength are related to foul, long, and strong, but they are not structurally related in the same way. Unlike the PGmc ancestors of filth, length, and strength, they were not derived from the corresponding adjective by addition of a suffix: instead, they were created by analogy with an existing pattern whose suffixal origin had been forgotten centuries earlier.
There are at least a couple more of these analogical creations. An early one is ME welthe ‘the condition of being happy or prosperous; riches, affluence’, the source of our wealth; it seems to have been formed in the later 1200s from ME wēle ‘worldly wealth; prosperity, good fortune; the common good’ (whence our weal), probably by analogy with health. A very early and very interesting one is ME slauþe, slouthe ‘indolence, slothfulness’, whence our sloth. Old English had an adjective sláw ‘sluggish, torpid; lazy; obtuse, dull’ and a matching noun slǽwþ ‘sloth, laziness’ that appear to go back to a PGmc pair *slaiwaz and *slaiwiþō. These became ME slāw, slōw ‘slothful, lazy’ (whence our slow) and early ME sleuthe ‘sloth, laziness’. By the mid-1100s, however, the latter was already starting to be replaced by ME slauþe, slouthe, which was formed directly from ME slāw, slōw.
I’ll finish this off with an oddball example that is not at all obvious and illustrates just how complicated this subject can be. The pair is sly and sleight (as in sleight of hand), which go back to ME sleigh ‘wise prudent; clever, ingenious; skilful, expert; crafty, cunning, deceitful’ and sleight ‘wisdom, prudence; cleverness, ingenuity; skill, expertness; strategy, planning; a clever stratagem, a trick’. These are borrowed from Old Norse slœgr ‘sly, cunning, crafty’ and slœgð ‘slyness, cunning’, which in turn are descended from Proto-Norse *slōgjaz and *slōgiþō, respectively. On the face of it this is an example of the PGmc suffixal formation, albeit one inherited in Old Norse rather than in Old English. However, there appears to be no evidence for it in any other Germanic language, and both words are probably very early North Germanic innovations. If so, the adjective was probably derived from the past tense stem slóg- of the verb slá ‘to strike’ and may originally have meant something like ‘able to strike’, and the noun was formed by addition of the suffix inherited from PGmc, which apparently was still productive. (The OE cognate of the verb was sléan ‘to strike, to beat; to kill by violence’, whence our verb slay.)