ons, because they make fad work when applied to their flimfy compofitions. But every man of true judgment, while he condemns diftinctions that have no foundation in the nature of things, muft perceive, that indifcriminately to decry diftinctions, is, to renounce all pretenfions to just reafoning; for as falfe reasoning commonly proceeds from confounding things that are different, fo without distinguishing fuch things, it is impoffible to avoid error, or detect fophiftry. The authority of Aquinas, or Suarez, or even of Ariftotle, can neither stamp a real value upon diftinctions of base metal, nor ought it to hinder the currency of those that have intrinfic value.

Some distinctions are verbal, others are real. The first kind diftinguish the various meanings of a word; fome of which may be proper, others metaphorical.Diftinctions of this kind make a part of the grammar of a language, and are often abfurd when translated into another language. Real diftinctions are equally good in all languages, and fuffer no hurt by tranflatior. They distinguish the different fpecies contained under fome general notion, or the different parts contained in one whole.

Many of Ariftotle's diftinctions are verbal merely ; and therefore more proper materials for a dictionary of the Greek language than for a philofophical treatise. At leaft they ought never to have been tranflated into other languages, when the idiom of the language will not juftify them for this is to adulterate the language, to introduce foreign idioms into it without neceffity or use, and to make it ambiguous where it was not. The dif tinctions in the end of Categories of the four words prius, fimul, motus, and habere, are all verbal.

The modes or fpecies of prius, according to Ariftotle, are five. One thing may be prior to another; first, in point of time; fecondly, in point of dignity; thirdly, in point of order and. fo forth. The modes of fimul are only three. It seems this word was not used in the Greek with fo great latitude as the other, although they are relative terms.

;

The

The modes or species of motion he makes to be fix, to wit, generation, corruption, increase, decrease, alteration, and change of place.

1. Hav

The modes or fpecies of having are eight. ing a quality or habit, as having wisdom. 2. Having quantity or magnitude. 3. Having things adjacent, as having a fword. 4. Having things as parts, as having hands or feet. 5. Having in a part or on a part, as having a ring on one's finger. 6. Containing, as a cafk is faid to have wine. 7. Poffeffing, as having lands or houses. 8. Having a wife.

Another diftinction of this kind is Ariftotle's diftinction of causes; of which he makes four kinds, efficient, material, formal, and final. Thefe diftinctions may deferve a place in a dictionary of the Greek language; but in English or Latin they adulterate the language. Yet fo fond were the schoolmen of diftinctions of this kind, that they added to Ariftotle's enumeration, an impulfive caufe, an exemplary caufe, and I do not know how many more. We seem to have adopted into English a final caufe; but it is merely a term of art, borrowed from the Peripatetic philofophy, without neceffity or ufe; for the English word end is as good as final caufe, though not so long nor fo learned.

SECT. 4 On Definitions.

It remains that we make fome remarks on Ariftotle's definitions, which have expofed him to much cenfure and ridicule. Yet I think it must be allowed, that in things which need definition, and admit of it, his definitions are commonly judicious and accurate ; and had he attempted to define fuch things only, his enemies had wanted great matter of triumph. I believe it may likewise be said in his favour, that until Locke's effay was wrote, there was nothing of importance delivered by philofophers with regard to definition, beyond what Ariftotle has faid upon that fubject.

He confiders a definition as a fpeech declaring what a thing is. Every thing effential to the thing defined, and nothing more, must be contained in the definition. Now the effence of a thing consists of these two parts:

Firft,

First, What is common to it with other things of the fame kind; and, fecondly, What distinguishes it from other things of the fame kind. The first is called the genus of the thing, the fecond its Specific difference.--The definition therefore confitts of these two parts. And for finding them, we must have recourse to the ten categories; in one or other of which every thing in nature is to be found. Each category is a genus, and is divided into fo many fpecies, which are diftinguished by their fpecific differences. Each of thefe fpecies is again fubdivided into fo many fpecies, with regard to which it is a genus. This divifion and subdivision continues until we come to the loweft fpecies, which can only be divided into individuals, diftinguifhed from one another, not by any fpecific difference, but by accidental differences of time, place, and other circumftances.

The category itfelf being the highest genus, is in no refpect a fpecies, and the loweft fpecies is in no refpect a genus; but every intermediate order is a genus compared with thofe that are below it, and a fpecies compared with thofe above it. To find the definition of any thing, therefore, you must take the genus which is immediately above its place in the category, and the fpecific difference, by which it is diftinguished from other fpecies of the fame genus. These two make a perfect definition. This I take to be the fubftance of Ariftotle's fyftem; and probably the fyftem of the Pythagorean school before Ariftotle, concerning defini

tion.

But notwithstanding the fpecious appearance of this fyftem, it has its defects. Not to repeat what was before faid, of the imperfection of the divifion of things into ten categories the fubdivifions of each category are no lefs imperfect. Ariftotle has given fome fubdivifions of a few of them; and as far as he goes, his followers pretty unanimoufly take the fame road. But when they attempt to go farther, they take very different roads. It is evident, that if the series of each category could be completed, and the divifion of things into categories could be made perfect, ftill the higheft

genus

genus in each category could not be defined, because it is not a fpecies; nor could individuals be defined, becaufe they have no fpecific difference. There are alfo many fpecies of things, whofe fpecific difference cannot be expreffed in language, even when it is evident to fenfe, or to the understanding. Thus, green, red, and blue, are very diftinct fpecies of colour; but who can exprefs in words wherein green differs from red or blue?

Without borrowing light from the ancient fyftem, we may perceive, that every definition must consist of words that need no definition; and that to define the common words of a language that have no ambiguity, is trifling; if it could be done; the only use of a definition being to give a clear and adequate conception of the meaning of a word.

The logicians indeed diftinguish between the definition of a word, and the definition of a thing; confidering the former as the mean office of a lexicographer, but the laft as the grand work of a philofopher. But what they have faid about the definition of a thing, if it has a meaning, is beyond my comprehenfion. All the rules of definition agree to the definition of a word: and if they mean by the definition of a thing, the giving an adequate conception of the nature and effence of any thing that exifts; this is impoffible, and is the vain boast of men unconfcious of the weakness of human understanding.

The works of God are all imperfectly known by us. We fee their outfide, or perhaps we difcover fome of their qualities and relations, by obfervation and experiment, affifted by reasoning; but we can give no definition of the meaneft of them which comprehends its real effence. It is juftly obferved by Locke, that nominal effences only, which are the creatures of our own minds, are perfectly comprehended by us, or can be properly defined and even of these there are many too fimple in their nature to admit of definition. When we cannot give precifion to our notions by a definition, inuft endeavour to do it by attentive reflection upon them, by obferving minutely their agreements and differences, and especially by a right understanding of the

;

we

powers

powers of our own minds by which fuch notions are formed.

The principles laid down by Locke with regard to definition, and with regard to the abuse of words, carry conviction along with them; and I take them to be one of the most important improvements made in logic fince the days of Ariftotle; not fo much because they enlarge our knowledge, as because they make us fenfible of our ignorance, and fhew that a great part of what fpeculative men have admired as profound philofophy, is only a darkening of knowledge by words without understanding.

If Ariftotle had understood thofe principles, many, of his definitions, which furnish matter of triumph to his enemies, had never feen the light; let us impute them to the times rather than to the man. The fublime Plato, it is faid, thought it neceffary to have the definition of a man, and could find none better than Animal implume bipes; upon which Diogenes fent to his fchool a cock with his feathers plucked off, defiring to know whether it was a man or not.

SECT. 5. On the Structure of Speech.

The few hints contained in the beginning of the book concerning Interpretation, relating to the ftructure of fpeech, have been left out on treatises of logic, as belonging rather to grammar; yet I apprehend this is a rich field of philofophical fpeculation. Language being the exprefs image of human thought, the analyfis of the one must correfpond to that of the other.Nouns adjective and fubftantive, verbs active and paffive, with their various moods, tenfes, and perfons, must be expreffive of a like variety in the modes of thought. Things which are diftinguished in all languages, fuch as fubftance and quality, action and paffion, caufe and effect, must be diftinguished by the natural powers of the human mind. The philofophy of grammar, and that of the human understanding, are more nearly allied than is commonly imagined.

The structure of language was purfued to a confiderable extent, by the ancient commentators upon this

VOL. III.

H

book