10* s. i. APRIL 9, 1904.] NOTES AND QUERIES.
291
N PRONOUNCED NG.
(10 th S. i. 247.)
I MUCH deprecate the discussion of phonetic questions. One who knows the answer is often placed in a false and unenviable position by being thus asked to explain technical matters which are properly treated in technical books, such as Sweet's 'History of English Sounds.' It requires preliminary knowledge, such as the majority do not possess, before an answer can be understood, unless one occupies far more space than can reasonably be given to the consideration of such a subject as this.
The very title assigned to the question shows how wholly the matter is misunder- stood. The true word is the spoken utterance ; the mere spelling is only the repre- sentation of such utterance, and often repre- sents it very badly. It is not the letter n that is pronounced as ng (though such vague expressions are only too common), but the sound of ng that is represented by n ; which is a very different way of putting it.
The fact is this. We have, in modern English spelling, adopted this rule, viz., always to represent the sound of ngk by the symbol nk. The rule has the convenience of saving a letter withoutcausing any ambiguity. For this reason it was that, even in Gothic, in which the symbol for the sound of ngk happened to be ggk (in imitation of Greek), it was not unusual to write gk simply ; hence the Gothic driggkan, to drink, was also written drigkan.
Similarly, instead of A.-S. dringcan, it seemed sufficient to write drincan. Wherever the symbol nc occurs in A.-S., it is to be understood as denoting the sound which would more correctly be denoted by ngk or ngc.
One great trouble is that ng denotes a simple elementary sound, and has, in philo- logical works, a special symbol. It is quite distinct from n followed by g. Neither the ng in sing nor the implied ngg in single is sounded like the ng in sun-god. This should always be borne in mind.
As the use of nk for ngk is invariable, no harm arises. But the sounds of ng^ in sing and in single, though quite distinct, are written alike. It may be well to show how this arose.
It simply arose from the fact that, at least in the earliest A.-S., and probably in the latest, the sound of ng in sing does not appear to have existed except before a consonant,
when its position decided its value. The A.-S.
sang, a song, was pronounced sangg (with
Italian short a), and singan, to sing, was pro-
nounced as singgan. But there came a time
when a final ngg was pronounced as ng simple,
giving a Middle-English sang or song, though
the verb remained as singgen. Then came a
time when the verb was reduced to sing-ge (two
syllables), then to singg, and then to sing. But
such reduction never occurred in words where
the sound of ngg was never final. That is why
we still say lingger and fingger and singgle,
whilst singer and songster are reduced to con-
formity with sing and song.
There is a great deal more to be said. I will only say, briefly, and (I hope) once for all, that no man can expect to have any real grasp of the principles of English spelling until he has learnt (1) the old Roman pronunciation of the Latin alphabet which we employ ; (2) the sounds and sound-laws of Anglo-Saxon ; (3) the sounds and sound-laws of Anglo-French ; and (4) the changes made by us both in sounds and symbols since A.D. 800. WALTER W. SKEAT.
The answer to W. S. B. H., who pertinently asks, " Why is the letter n sounded as ng before k, &c.1" is that the practice is chiefly an out- come of the loose and careless way of speak- ing which has long since spoilt some of our habitual locutions. It is certainly worse in my own recollection. But those persons who have a mind to preserve the more cultivated phases of the English tongue will continue to say an-chor, an-guish, Jenkins, and so forth.
After purchasing Annandale's ' Concise English Dictionary' I was amazed to find these "pronunciations" given: anchor= angker; ankle = angkl ; an ky losis = ang-kilosis; but encroach=en-kroch ; enquire==en-kwir ; also incqnvenient=inkonvenient ; increase^ inkres ; inquire=inkwir, &c. It looks " ing- konsistent" to treat an differently from en and in. One cannot find refuge in respect to the accent that is to say, apply the g to the prefix when it is accented. No; it is neither more nor less than needless haste in speaking, and consequent failure to be elegant.
EDWARD SMITH.
It is, I think, convenience of speech alone that dictates the ringing sound of ng in words where the semi-vowel n precedes a k. If, in ignoring any g sound, one were to repeat a dozen times any one of the words, such as "anchor," that W. S. B. H. has named, it would be found that more time and trouble would be necessary than would be involved in the articulation of the g sound. Custom and convenience make the pronunciation
