Most verbs in English work like this, adding an “s” in the case of the third-person singular:
- I sing. You sing. He sings.
- I walk. You walk. She walks.
A few verbs in English work like this, keeping the same form for all present tense forms:
- I can. You can. He can.
- I will. You will. She will.
But there is one special verb that is highly irregular:
- I am. You are. He is.
What’s going on here? In particular, let’s focus here on this question: Where does “am” come from?
It turns out that “am” is the last vestige in English of an ancient verb conjugation shared by Sanskrit, Greek and Latin. In the middle of Eurasia around 4500–6500 years ago., there was a tribe who spoke a language in which the equivalent of “am” was *esmi. As the descendants of this tribe spread out in different directions over thousands of years, its language evolved into hundreds of different languages from India to Europe, known collectively as the Indo-European languages. Using linguistic evidence, the original tribe’s language has been partially reconstructed, and is referred to as Proto-Indo-European (PIE).
Let’s take a look at how the verb *esmi from Proto-Indo-European evolved in some early Indo-European languages:
- esmi (Hittite, part of the extinct Anatolian family)
- ásmi (Sanskrit)
- jesmĭ (Old Church Slavonic, the oldest written Slavic language)
- sum (Latin)
- eimi (Homeric Greek)
- aʰmiy (Old Persian)
- *immi (Proto-Germanic)
It’s easy to see the similarity of *esmi (PIE) with its forms in Hittite, Sanskrit and Old Church Slavonic.
In Latin, where *esmi (PIE) evolved into sum, the vowels changed a lot, but the consonants remained the same. This verb sum famously appears in Descartes’ philosophical aphorism cogito ergo sum (I think, therefore I am).
In the ancient Greek, Persian and Germanic languages, the “s” sound disappeared, leaving an “m” between two vowel sounds. The Proto-Germanic *immi became im in Gothic, em in Old Norse, and eom in Old English — which in turn became “am” in modern English.
Note that the -mi ending of *esmi was also used for many other verbs in Proto-Indo-European to indicate the first-person singular for the present tense. This widespread usage of -mi was preserved in a few descendant languages:
- In Sanskrit, this -mi ending became used for all verbs. For example, in Sanskrit, bhavāmi means “I become” and karomi means “I make” in Sanskrit.
- In Persian (also known as Farsi), *esmi became am (the same as in English!), and the ending -am became used for all present tense verbs in the case of the first-person singular. For example, man mikonam means “I do” and man mixoram means “I eat.”
- In Polish, *esmi became jestem, and there are 15 common irregular verbs that use -m for the first-person singular. For example, mam (“I have”) and mieszkam (“I live”).
In most other Indo-European languages, however, this -mi conjugation from Proto-Indo-European has only been preserved in the descendants of *esmi:
- sum in Latin, eimi in ancient Greek and “I am” in English
In contrast, for all other verbs, there’s no “m” in the first-person conjugation. For example:
- scribo in Latin, grapho in Greek, and “I write” in English.
It’s not surprising that the verb “to be” in English is highly irregular — in language evolution, the most commonly used words tend to change more slowly and retain irregularities. As “I am” is one of the most basic phrases in English, this verb has kept its irregular form over thousands of years, preserving a link to an ancient conjugation that was used for all verbs in Sanskrit and remains widely used (in a variant form) in Persian today.
Notes:
- The asterisk (such as in *esmi) indicates a hypothesized reconstruction.
- Note that linguists also believe that there may have been some additional laryngeal sounds in *esmi, but I’ve omitted them for simplicity.
- In the Romance languages, the Latin sum lost its “m” sound to become soy in Spanish, suis in French, and sou in Portuguese. The “m” became an “n” sound to become sono in Italian.
References:
