In part 1 we looked at grammatical acceptability judgements for variations on the sentence “Me and my brother were chatting with my neighbor”. We observed two tendencies:

• “I” is generally preferred in the subject of the sentence, and “me” is preferred in the object of “with”.
• Overall, “I”/”me” sounds better after “my brother” than before, and “I” is preferred over “me” after “my brother”, while “me” prefers to come before “my brother”. Also, switching the order of “I” and “my brother” affects things more than switching the order of “me” and “my brother”.

We saw, however, that “I and my brother were chatting with my neighbor” and “My neighbor was chatting with my brother and me” are less acceptable than these two generalizations would predict. Furthermore, some of the other sentences seem to be bit more acceptable than a simple application of these rules would predict.

So what else is going on? And really, why do people show these acceptability preferences?

Here, for your enjoyment, are a bunch of numbers. As has been discussed extensively on LanguageLog and Corpora List, hit counts can be fairly unreliable, but they do give some indication of frequency of usage.

(BNCWeb is 100M words of British English. Some searches pull up hits for other constructions, and in those cells the number in parentheses is the actual hit count, and the other number is my estimate of the portion of those hits that actually represent the language usage we’re looking for. I’ve switched to present tense here (”I am“), where there are more distinct conjugations of “be”, to help reduce these undesired hits.)

If you pull up part 1 in a new tab and compare the two tables, you’ll notice that within the groups (b-e, f-i, j-m) there’s a fair correlation between high acceptability ratings and high hit counts. The sentences judged 1 barely have any hits (or none), ranging up to the sentences judged 5 having millions of hits. One exception is that “with my brother and me” scored fairly high in acceptability, but fairly low in usage. Apparently either this construction is more common among college-age Californians than among other English speakers, or it’s a form that they accept, but don’t actually produce. As I explain below, I think it’s this second option.

Oddly enough, however, all the constructions except “with I” have some usage. (And actually, if we had the patience, we could probably find some real examples among the millions of “with I” hits.) Take for example, “me am” and “with I and my brother”. Even these constructions rated really unacceptable have people using them, often enough on purpose. Y R THEY BUCHERING TEH BUTIFUL ENGLISH LANGUGE?!!1! Why? Because even officially incorrect language forms have contexts in which they are appropriate, even more appropriate than the “standard” forms. Livejournalist Illusion saying “me am happy and now shall do the happy dance” expresses something different than “I am happy…” In using the non-standard language she is expressing an abnormal, goofy mood.

Let’s return now to the puzzle. You may have already noticed that you would be more likely to use these constructions in some contexts than others. If your own intuitions haven’t already convinced you, compare the contexts for “Me and my brother are” with the contexts for “My brother and I are”, and look at how the count ratio changes when you add other words like “alas” (ratio goes from 1:4 to 1:9)[1][2] or “omg” (ratio goes from 1:4 to 1:2)[1][2]. There may also be a dialect difference, but there is at least a register difference, i.e. a difference in context of use.

If we take another look at the regularities we found earlier, keeping in mind that the favorite forms (”my brother and I” and “me and my brother”) represent slightly different language varieties, we can see a rather interesting situation. While the single pronoun differs between subject and object, and it does not differ between register, the compound noun phrase differs between registers, and it does not differ between subject and object!

Those of you who had a fair amount of grammar in school might be all like, “Wait! That’s not how academic English works.” And you’d be kinda right. Kinda. “With my brother and me” is the traditionally grammatical form. It was rated as fairly acceptable, but not often used, probably because we’re exposed to it as an officially grammatical form, but disagrees with our experience, which is ultimately what our productive grammar is derived from. “With my brother and I” is hard to count because so many of the hits are something else (like “with my brother, and I …”), but it appears that when “with my brother and me” is used, it does tend to be a bit higher social context than “with my brother and I”. The fact is, while “with my brother and me” has been maintained in teaching, “with my brother and I” is more common usage except in the most carefully edited language, and has been around for a long time.

So what all this illustrates is that since language is constantly changing, and a single language has different varieties appropriate to different contexts, as well as varieties spoken by different people, the common concepts of “bad grammar” and “proper English” are not very coherent.

Some may say that “my brother and I were” and “with my brother and me” are inherently more proper because they are more logical, in that they maintain the rules that “I” goes in the subject and “me” in the object, and you always put others before yourself. But the fact is, language isn’t completely logical. While human language does have many statistical regularities, and even some rules that seem to always be followed, real human language has all kinds of idiosyncracies and exceptions. The processing in the human brain is analog, associative and probabilistic, not symbolic, deductive or deterministic. The language that results from that processing has organic, chaotic order, not the order we get from formal logic.

None of the constructions above is the most appropriate option in all contexts, and most of the constructions have social contexts in which they are appropriate options. It may be “improper” to write “Me and my brother was …” in a resume cover letter, but equally improper for a cholo to say to his cuates, “My brother and I were …” The only sense in which there is proper language is in that for every social group and social context, there is set of acceptable language forms, and in order to be accepted in a social group, you have to be able to produce the appropriate language. Everybody has to be able to expertly alternate between language varieties as they change social contexts. As the phrase is most often used, learning to use “proper English” is really about learning which language forms belong in which contexts, and since no one grows up speaking with the same forms used in formal written English, all of us, some more than others, have to learn this new language variety that the general society has constructed and given high social status to, to be accepted in certain scholastic, employment and civic contexts.

The trouble comes in that we confuse high social status with inherent quality. We associate it with education, since it does take education to learn formal English, but we ignore the fact that some home language varieties are more like formal English than others, and that some people’s education focuses more on formal English than others. Perhaps because of education, we associate formal English with intelligence, ignoring that it is just as much if not further removed from the observed behavior. And somehow there is a pernicious tendency to see lower status language varieties as less logical and motivated by laziness. The result is that we fail to appreciate all the other language varieties. And that’s kinda sad.