Why do we say “I am a teacher” instead of “I is a teacher” considering that I is a singular pronoun not a plural pronoun? Don’t singulars always take -s forms? Why does be work differently?
Whilst this is indubitably an odd question, it is not as hollow or uninteresting as it sounds at first blush. Grammatical number generally coincides with actual number. It is a concrete feature of the entity or entities denoted. Verbs in many languages agree in number with their subjects. The types of number that they agree with vary, however. So some nouns in some languages can inflect for being singular, dual (indicating there being two of them), and nondual plural number. However, generally speaking, grammatical number is a feature that co-incides with facts about the denoted entities.
Contrastingly, what we call grammatical person is not a feature of any entities at all. We can try to ascribe features to the first or second person, for example, but these fancies will be doomed to failure. We can say, for example, that the first person is self-ascribed to the person who is speaking - but only whilst they are actually speaking. However, this is nonsense. To illustrate, I, Araucaria, have the status of 'speaker' as you are reading this. Now it is only if I choose to refer to myself using a pronoun that the verb will agree with my 'speaker status':
- *Araucaria am addressing you right now!
- Araucaria is addressing you right now!
So no objective attribute in terms of who is speaking or being addressed is enough - in English - to cause subject verb agreement in this way. The same phenomenon occurs with the second person and you:
- *The reader are reading this sentence.
- You are reading this sentence.
Grammatical person in English therefore seems not only to be deictic - a kind of pointy feature of certain words - and therefore not an actual feature of any particular entities being denoted, but also a feature that is only reflected by verbs in relation to the subject being - or not being - a certain type of pronoun.
For the most part, verbs in English only inflect for person in the third person singular in the present tense. In fact outside of the verb BE, no other verb in English inflects for person in any other case. Were it not for the verb BE it would be difficult to actually decide whether verbs in English really inflected to agree with grammatical person at all. We could for example say they inflect to agree with having the feature gender, because only third person singular forms in English have gender - no other ones do. The verb BE however, puts a spanner in the works of this theory.
This brings us back to the Original Poster's question of why the verb BE inflects to reflect first person - and arguably only in the present tense. There is another directly related issue, namely that, very confusingly, in negative interrogatives I, in fact, does appear to take plural verb agreement:
- Aren't I?
For more on that anomaly see Why "ain't I" and "aren't I" instead of "amn't I"?. The Original Poster doesn't, rather puzzlingly, ask why you seems to take what might be thought of as a plural verb form agreement - which might seem to be a more pertinent question. BE seems to be a very strange verb.
Why BE should be so weird is a different issue. Aside from irregularities stemming from pronunciation constraints, irregular verbs in English, and indeed in many other languages seem to be high frequency words. BE obviously qualifies in this respect.
The verb BE is indeed irregular in many languages, and furthermore, many languages do not have an equivalent verb at all. The verb BE is usually semantically empty and contributes nothing in terms of meaning to the sentence. Sentences like:
- Bob is happy
... would literally have no verb in many other languages. That sentence would read as Bob happy, for example, if translated word for word from Turkish.
We can see then that BE has an unusual function and is unusual semantically. I often, in fact, tell my students that there is no main verb BE in English at all. The reason for this is that, grammatically, BE is an auxiliary verb. It contracts with not, inverts with subject to form questions and so on and so forth. We can then represent the following sentences thus, where each sentence has a subject, auxiliary verb and lexical verb position:
- She [can] [dance].
- She [has] [finished].
- She [Ø aux] [smokes].
- She [is] [Ø verb] happy.
This structure can be substantiated by putting in an adverb such as usually:
- She [can] usually [dance].
- She [has] usually [finished].
- She [ Ø ] usually [smokes].
- She [is] usually [ Ø ] happy.
We can see that usually now occurs in the same post auxiliary position in each sentence.
So BE in English apart from having no meaning, being extremely high frequency, and behaving as an auxiliary, is even defective amongst auxiliaries in being able to occur without a main verb!
Given all of these anomalies surrounding BE, it is really not so surprising that it inflects for grammatical person where other verbs don't, and even has a different inflection for the first person depending on whether it is part of a negative interrogative construction. Now this doesn't really answer the Original Poster's question as such, but hopefully shows why it might be an interesting question of sorts.
The conjugation of to be in present tense is
I am, you are, he/she/it is - plural: we/you/they are.
Questions as why can't I say I are or we am don't make much sense. This is the system the speakers of English have and you have to accept it.
Historical explanations of why verbs have different forms for different persons are not easy to give. But I will try to give you my view taking Latin as an example:
In Latin the verb for to love has the forms (present tense)
'am-o 'am-as 'am-at - plural: a'ma-mus a'ma-tis 'am-ant.
The endings of the persons are o as at - amus atis ant. Stress marked with ' is alway on the second syllable from behind.
Pupils learn these ending mechanically, nobody tells them something about the logic of these endings. It is this way, point.
I was already an adult man when I had the idea that these endings must be remainders of the personal pronouns. This is easiest to see in the form amamus we love. The Latin personal pronoun for we is nos. So we may assume that the ending -mus was originally nos and that Latin actually said "love-we" or "loving-we" or "in love/in loving-we". In any case the personal pronoun was added after a certain verb form as a kind of suffix, and, of course, to ease pronunciation, was slightly changed and often shortened. This system can also be seen easily in the form amo, I love, or "love+I". The ending -o must be the remainder of the personal pronoun ego meaning I.
For the other forms the formation is more difficult, but I think, you understand the system.
The system with person-endings has one disadvantage. The endings melt with the verb and irregular forms arise. Modern languages place the personal pronouns before the verb so that there is no melting process. But as we see in Italian and French, even in German, the historical endings were not simply dropped. So we have the following systems:
Latin amo amas amat - amamus amatis amant
Italian amo ami ama - amiamo amate amano
Italian can use personal pronouns before the verb forms, but it is not necessary.
French j'aime tu aimes il aime - nous aimons vous aimez ils aiment
In French the endings are reduced as your hear only the endings ons et ez.
German ich liebe du liebst er liebt - wir lieben ihr liebt sie lieben
German still has four endings.
English has kept only one ending:
I/you love he loves - we/you/they love.
The verb to be is the only irregular verb that has three forms am are is. Actually the forms of the present tense of to be are derived from three different verbs.
I hope you see that such systems with personal pronouns and verb endings have evolved in periods of time of two thousand years and I hope you will understand that you can't change this system arbitrarily as you like.