The word grammar is often used to mean "formal rules for correct writing." In this sense, grammar is a slightly artificial subject, and rules of grammar are laid down by grammatical authorities who decide what is and what isn't good English (or French, or Spanish, or Russian, or whatever). However, there is another sense of grammar that philosophers in particular should be aware of. Quite apart from any formal instruction, anyone who speaks a language has a command of an enormous set of rules that determine what the sentences in that language are. Children learn most of the grammar of their native language, in this sense, by the time they are about five, and they learn it just by being exposed to people who speak the language.
These two senses of "grammar" are related, of course. The grammar taught in school is based on the grammar naturally learned by children and implicitly understood by native speakers. It's common to call the kind of grammar taught in schools prescriptive grammar. The remarks here are mostly about prescriptive grammar.
It's important to remember two things about prescriptive grammar. First, it contains many rules that are rather arbitrary, and in fact you can encounter rules that are unreasonable or silly. Second, whether you follow its rules can affect how your writing is perceived by other people. This is perhaps not an entirely desirable state of affairs. Some people have a very realistic (in the philosophical sense) view of grammatical standards: they are quite sure that there is a fact of the matter about what's correct, and they may regard their view of what's correct as determined by logic or rationality. Even if some rules in common circulation have a shaky basis, however, you still need to be aware that they exist. It will save you some embarrassment and annoyance.
In order to understand traditional grammar, it's necessary to be familiar with a certain amount of terminology for analyzing sentences. The terminology is actually useful for other purposes as well, including the study of logic and the philosophy of language, so it won't hurt to spend a little time on it.
A sentence, grammatically speaking, is at least one clause. A clause is, roughly, a subject and a verb, with other modifiers possibly attached to them. Here are some simple one-clause sentences:
Sentences like these, made of single clauses, are called simple sentences. The clauses of which they are composed can be called independent clauses because they can stand alone as complete sentences. Clauses like these can be combined into compound sentences by the use of what are called coordinating conjunctions. Warning: the word "conjunction" is used in quite different ways by grammarians and by logicians. For a logician, a conjunction is a truth-functional compound sentence with the property that it is true if and only if all of its constituents are true. For our present purposes, a conjunction is a word that combines simple sentences into compound sentences. Three very good examples are "but", "or", and "and":
There are also clauses that can't properly stand on their own as independent sentences. These sentences contain some examples (in boldface):
These examples include very different sorts of dependent clauses. One thing to be aware of from the start is that logicians and grammarians may look at sentences very differently. Notice in particular the last example. A logician will analyze this sentence as a conditional, with two component sentences "Nominalism is correct" and "Numbers do not exist" joined by the connective "if". A grammarian will instead see it as an independent clause "Numbers do not exist" and a dependent clause "If nominalism is correct". However, logician and grammarian would concur on reading "Fluffy is a dog but Rover is a cat" as combining two independent clauses with the connective (or conjunction, from the grammarian's point of view) "and".
Logical analysis and grammatical analysis may also diverge when it comes to sentences with compound subjects and compound predicates. Consider these two examples:
If you've had a class in symbolic logic, you have probably learned to translated the first sentence as an abbreviated way of expressing the second. Each is a conjunction (logically speaking) of "Leibniz founded symbolic logic" and "Frege founded symbolic logic". However, from a grammarian's point of view the first sentence is a simple sentence with the compound subject "Leibniz or Frege", while the second is a compound sentence. This affects good writing in a number of ways. For one trivial example, there is a comma in the second sentence. It is usually a good idea to separate the clauses of a compound sentence with a comma before the conjunction, unless the sentence is very short. However, there's no place for a comma in the first sentence.
In standard English, two separate negations in the same sentence operate independently. Thus, "I didn't do nothing" should mean "It is not the case that I did nothing" (which is logically equivalent to "I did something", though not really the same sentence). However, in colloquial (some would say illiterate) English, two negatives can sometimes reinforce one another: "I didn't do nothing" is a more intense way of saying "I did nothing" or "I didn't do anything". Formal grammarians have ruled that the latter kind of double negative is 'illogical' and thus not to be allowed (although Shakespeare is not above using them, and although in French, Spanish, and many other languages two-part negations are actually required). Whatever the merits of traditional grammar on this score, it's best to avoid sentences containing double negatives (like "I didn't do nothing") regardless of which way you mean them.
In English, agreement is for the most part a matter of having singular subjects with singular verbs and plural subjects with plural verbs. In the sentence "Utilitarians believe that the action with the greatest overall benefit is right", the verb is "believe" because that is the third-person plural form and "utilitarians", its subject, is plural. In "A utilitarian believes that the action with the greatest overall benefit is right", the verb is the singular form "believes" because the subject, "a utilitarian", is singular. In English, there is not a whole lot more to agreement than this (you do need to watch the verb "be", which has a separate form "am" for the first-person singular). In complicated sentences, however, it is sometimes easy to lose track of the subject and consequently to make mistakes in agreement.
One problem area for agreement is words that function as (speaking in the language of logic) quantifiers. Some of these are grammatically singular, some are grammatically plural, and some can be either singular or plural. "Each", "any", and "every" should be treated as singulars; "all" should be treated as plural, and "some" and "no" can go either way. Consider these examples:
A somewhat more vexing problem is pronoun agreement. English has simplified its grammar considerably over the last 600 years or so, but we still have personal pronouns ("I", "we", "you", "he", "she", "it", "they") that change in person ("I" vs. "you") and case ("she" vs. "her", "they" vs. "them") as well as number (singular or plural). Pronouns come in three cases:
For the most part, native speakers of a language make correct pronoun choices without a lot of effort. However, matters can get complicated with compound subjects. One reason for this is that English has just enough grammatical gender left in it to be an irritant. Consider these examples:
The problem that some of these examples illustrate is what to do about cases where we need a singular pronoun (he, she, him, her) or adjective (his, her) but don't know whether the person in question is male or female. The old-fashioned solution, taught a couple of generations ago, was to use masculine pronouns exclusively in such cases. However, this is now widely perceived as sexist, that is, as making the assumption that people are all male if not otherwise specified. The effect of this is clearer from examples such as this:
Does this example assume, or at least suggest, that every philosopher is a man rather than a woman? In order to avoid this problem, three solutions have been adopted. One is to use plurals ("they", "their", "them") with singular verbs when the gender of the person is unknown or undetermined:
In spoken English, this is actually the solution almost everyone has been using for centuries. However, in some cases it sounds a little awkward:
A second solution is to use constructions like "he or she", "her or him", "her or his". These work, though they are perceived by some writers as clumsy:
A third solution used by some writers is to alternate between feminine and masculine pronouns/adjectives in different cases.
My personal opinion is that it's best to avoid the need for these solutions, where possible, by switching to the plural:
And, where that's not possible, use "they/them/their" (writers have actually been doing this for centuries). It is worth bearing in mind that this is a problem in English only because of details about the way English grammar works. Other languages may avoid these problems, or they may have different problems. For example, in Romance languages (Spanish, French, Italian, Portuguese, etc.), all nouns, as well as the words for "the" and "a", have either masculine or feminine gender: it is not possible to say "a philosopher" without making the "philosopher" masculine or feminine. On the other hand, in these same languages, the words for "his" and "her" are the same (actually, they do have a gender, but it's grammatical gender that agrees with the word they modify). In still other languages (e.g. German and Russian), there are three grammatical genders, with resultant complications. And in still others (e.g. Finnish), there is no grammatical gender at all, eliminating these problems. One lesson that might be extracted from all this is that philosophers would do very well to learn more than just one language.
In English, an infinitive has the form "to" plus a verb: "to go", "to see", "to argue". Infinitives also have passive forms ("to be seen") and past forms ("to have gone"). Older grammatical rules say that it is wrong to "split an infinitive", that is, to insert anything between "to" and the verb: "to boldly go", "to willingly injure", "to deliberately lie". As a matter of fact, many good writers are not all that fussy about splitting infinitives. Historically, this rule was devised on the basis of Latin grammar, which is somewhat bizarre since in Latin, infinitives are (with few exceptions) single words. However, even though it's arbitrary, it usually makes for a clearer style of writing if you avoid putting things in between "to" and its verb when you can do so.
English spelling is just plain complicated, mostly for historical reasons. Once upon a time, letters had fairly consistent sound values, and people could spell things rather on the basis of how they sounded. You can still do that in many European languages, but English pronunciation started drifting away from spelling a few centuries ago, and unfortunately we've hung on to spellings that no longer have much to do with the way the word is pronounced today (consider "knight", where the "k" and the "gh" once had sounds but no longer do, or "one", or all the different sounds of "ough" in "rough", "ought", "bough", "though", "through"). You just have to learn how to spell.
Spell-checkers in word processors can help, but unfortunately English is also full of homonyms (words that are pronounced alike but spelled differently), and a spell checker will only tell you that the word you wrote really is an English word, not whether it is the word you intend to be writing: "to", "too", and "two" are all English words, as are "there", "their", and "they're". Here are some things to watch:
There are also near-homonyms that are sometimes confused: "then" and "than".
Some words borrowed from Greek or Latin are commonly pluralized as they would be in those languages. In the opinion of this particular writer, this shouldn't be overdone: if the Latin or Greek plural is in standard use, then use it, but watch out about providing these if you're not sure. You may in fact get it wrong.
The most common Latin plurals are of words ending in -us, with plurals in -i, or words ending in -um, with plurals in -a (datum, data). Many Latin words ending in -a in the singular have plurals in -ae (differentia, differentiae). Many words of Greek origin ending in -on have a plural in -a (criterion, criteria; phenomenon, phenomena).
Unfortunately, there are other cases where these rules don't work. "Schema" is a Greek word, but its Greek plural is "schemata", not "schemae". The Latin plural of "corpus" is "corpora", not "corpi." And the Latin plural of "species" is "species" (in English as well as Latin, "species" is both singular and plural, like "deer" and "sheep": avoid saying or writing "specie"). "Genus" pluralizes to "genera". "Virus" is a particularly weird case, since it was never actually used in the plural in Latin and has what is called a "heteroclitic" declension (just say "viruses" and you're always right).
There just isn't any simple rule to tell you what the plurals of such words are: you have to learn them (or learn Latin and Greek). At any rate, avoid the assumption that any word that ends in -us should have a plural in -i.
Here's a table of a few examples (the preferred form, according to most dictionaries, is in boldface):
| Singular | Fancy Plural | Plain Plural | Wrong Plural | Wrong Singular | Notes |
|---|---|---|---|---|---|
| genus | genera | ||||
| differentia | differentiae | differentium | |||
| schema | schemata | schemas | schemae | ||
| curriculum | curricula | curriculums | |||
| bacterium | bacteria | bacteria | That's right, saying "a bacteria" is incorrect, but very widespread | ||
| opus | opera | Don't confuse the plural with a type of musical composition | |||
| phenomenon | phenomena | phenomenon | phenomena | Likewise for "criterion" and lots of other words: "a criteria" is wrong. | |
| datum | data | ||||
| species | species | specie | the phrase "in specie" means "in coin" (i.e. metal money), so "specie" sometimes means "coins". For the record, "specie" here is the ablative singular of "species". | ||
| medium | media | mediums | Which plural is preferred depends on the meaning. If you must say "a media", please do so where I can't hear you. | ||
| octopus | octopi | octopuses | The Greek plural of this Greek word would actually be "octopodes" | ||
| syllabus | syllabi | syllabuses | ("syllabuses" is the preferred plural in British use) | ||
| virus | viruses | viri, virii | In Latin, "virus" ("poison") has no plural; "viri" would be the Latin plural of "vir", which means "man"; "virii" is just dumb | ||
| nexus | nexuses | nexi | (The Latin plural is actually "nexus", which nobody uses in English) | ||
| ethics | ethics | This and other words for academic fields or subjects, e.g. metaphysics, physics, mathematics, are grammatically singular: "mathematics is hard", not "mathematics are hard". You do sometimes find the singular forms ethic, metaphysic in older usage, but I suggest you avoid them. Watch out for physic, which has a totally different and potentially embarrassing meaning. |
There's a related class of plurals that originate as jokes but sometimes become widely used: plurals deliberately formed as if the word were, say, an Old English or a Latin word. Examples of jokes: kleenex, kleenices; sheriff, sheriffim (phony Hebrew plural thanks to Ogden Nash). Computer people are especially likely to come up with these (shouldn't the plural of "Unix" be "Unices"?), and in that field they even become standard. These are fine in their context, but unless you're sure they should be avoided. Safest rule to follow: pluralize as if in English unless you actually know that another form is correct.