that is countenanced by her, to be esteemed good, and therefore worthy to be retained?' SECTION I. Good use not always uniform in her decisions. IN answer to the former of these questions, I acknowledge, that in every case there is not a perfect. uniformity in the determinations, even of such use as may justly be denominated good. Wherever a considerable number of authorities can be produced in support of two different, though resembling modes of expression for the same thing, there is always a divided use, and one cannot be said to speak barbarously, or to oppose the usage of the language, who conforms to either side *. This divided use hath place sometimes in single words, sometimes in * The words nowise, noway, and noways, afford a proper instance of this divided use. Yet our learned and ingenious lexicographer hath denominated all those who either write or pronounce the word noways, ignorant barbarians. These ignorant barbarians (but he hath surely not adverted to this circumstance) are only Pope, and Swift, and Addison, and Locke, and several others of our most celebrated writers. This censure is the more astonishing, that even in this form which he has thought fit to repudiate, the meaning assigned to it is strictly donformable to that which etymology, according to his own explication, would suggest. See Johnson's Dictionary on the words nowise and way, particularly the senses of way, marked with these numbers, 15, 16, 18, and 19.. construction, and sometimes in arrangement. In all such cases there is scope for choice; and it belongs, without question, to the critical art, to lay down the principles, by which, in doubtful cases, our choice should be directed. There are, indeed, some differences in single words, which ought still to be retained. They are a kind of synonymas, and afford a little variety, without occasioning any inconvenience whatever *. In arrangement too, it certainly holds, that various manners suit various styles, as various styles suit various subjects, and various sorts of composition. For this reason, unless when sonie obscurity, ambiguity, or inelegance is created, no disposition of words which hath obtained the public approbation, ought to be altogether rejected. In construction the case is somewhat different. Purity, perspicuity, and elegance generally require, that in this there be the strictest uniformity. Yet differences here are not only allowable, but even convenient when attended with correspondent differences in the application. Thus the verb to found, when used literally, is more properly followed by the preposition on, as, The "house was founded on a rock;' in the metaphorical * Such are, subterranean and subterraneous, homogeneal and homogeneous, authentic and authentical, isle and island, mount and mountain, clime and climate, near and nigh, betwixt and between, amongst and among, amidst and amid. Nor do I see any hurt that would ensue from adding nowise and noway to the number. application, it is often better with in, as in this sentence, They maintained, that dominion is founded in grace.' Both sentences would be badly expressed, if these prepositions were transposed, though there are perhaps cases wherein either would be good. In those instances, therefore, of divided use, which give scope for option, the following canons are humbly proposed, in order to assist us in assigning the preference. Let it, in the mean time, be remembered, as a point always presupposed, that the authorities on the opposite sides are equal, or nearly so. When those of one side greatly preponderate, it is in vain to oppose the prevailing usage. Custom, when wavering, may be swayed, but when reluctant will not be forced. And in this department a person never effects so little, as when he attempts too much *. * For this reason it is to no purpose with Johnson to pronounce the word news a plural, (whatever it might have been in the days of Sidney and Raleigh) since custom hath evidently determined otherwise. Nor is the observation on the letter [s] in his Dictionary well founded, that "it seems to be established as a rule, that no noun singu<lar should end with [s] single;" the words alms, amends, summons, sous, genus, species, genius, chorus, and several others, show the contrary. For the same reason the words averse and aversion, are more properly construed with to than with from. The examples in favour of the latter preposition, are beyond comparison outnumbered by those in favour of the former. The argument from etymology is here of no value, being taken from the use of another language. If by the same rule we were to regulate all nouns and verbs of Latin construction, and sometimes in arrangement. In all such cases there is scope for choice; and it belongs, without question, to the critical art, to lay down the principles, by which, in doubtful cases, our choice should be directed. There are, indeed, some differences in single words, which ought still to be retained. They are a kind of synonymas, and afford a little variety, without occasioning any inconvenience whatever *. In arrangement too, it certainly holds, that various manners suit various styles, as various styles suit various subjects, and various sorts of composition. For this reason, unless when sonie obscurity, ambiguity, or inelegance is created, no disposition of words which hath obtained the public approbation, ought to be altogether rejected. In construction the case is somewhat different. Purity, perspicuity, and elegance generally require, that in strictest uniformity. Yet differe only allowable, but even conve with correspondent differenc Thus the verb to found, wh properly followed by the house was fou Beas application, is often better with in, as sentence founded in pre maintained, that dominier Both sentences would be the lem propreci expressed, f inse prepositions were tran of late djective il extem , and, he é. I know mention the ld make one eculiar to Mr on in good use searching back in good use to e ought to prefer which is an adjecdverb, to exceeding, same reason also I ,which makes ye inal of the personal prousative, when applied to used for the singular termined that it shall be renders the distinction rafor the most part it would www |