by Christina Skelton
Everyone who has studied Latin has wondered
about the verb “to be.” Why is there
all the irregularity? Why do we see es- and
s- in the present tense as in sum, es, est, but
er- in the imperfect, as in eram, eras, erat?
It seems to defy explanation! However,
from such seemingly irregular paradigms,
historical linguists can detect fascinating
clues to the past history of a language.
All languages evolve over time. Pronunciation
is just one of the many ways in
which languages can change, as one notes
just by listening to different dialects of
American English. One method of sound
change is unconditioned sound change.
For example, the words cot and caught are
pronounced with different vowel sounds
by some speakers, but identically by others.
These two different vowel sounds were
originally present in the English language.
However, they have come to be pronounced
as the same vowel sound in every
word in which they occur, at least for speakers
whose dialect has that sound change.
Likewise, for some speakers of American
English, the words pen and pin are pronounced
with two different vowel sounds,
while in others they are pronounced identically.
In this case, however, the merger of i
and e has only occurred before nasal consonants.
This is an example of a so-called
conditioned sound change, because the
change has only occurred within a specific
environment. Sound changes take place in
every word in the language in which the
sound and its conditioning environment
occur. Sound changes which involve the
merging of two sounds are also irreversible.
Once the sound change has taken place,
speakers do not know which of the sounds
in question a given word originally had.
Linguists, however, can search a language
for clues that a sound change has
occurred in its earlier history. We can look
for irregularities in the language which
could be explained if we assume that the
language was regular at an earlier stage,
but a sound change has introduced the
irregularity we see. In order to do this, we
first need to hypothesize an earlier, regular
form of an irregular stem. Then we need to
propose a sound change which, when applied
to the earlier forms, will produce the
irregular forms that we see.
Let’s look again at the paradigm of the
Latin verb “to be”:
Present Imperfect
sum sumus eram eramus
es estis eras eratis
est sunt erat erant
The stem of the verb “to be” appears to
alternate between s-, es-, and er-. Which
one is original? We can decide between
them if we remember the fact that sound
changes do not happen at random. Instead,
they only occur in a specific environment,
one consisting of the sounds which must
appear before or after the sound in question
in order for the sound change to take
place. We can look at each possibility in
turn, and then see which one would let
us propose a sound change with a specific
conditioning environment. So the question
is whether the verb “to be” originally had a
stem in s or r.
What happens if we assume that the
stem of the verb “to be” originally ended
in r? In that case, we are left with a wildly
varying set of conditioning environments
in which the sound could have changed
to s. An e may or may not occur before
the s, and the s may occur before the end
of the word, or before t or u. On the other
hand, if we assume that the stem originally
ended in s, we see a consistent conditioning
environment: we find r after e, and
before a. The stem of the verb “to be” must
have originally ended in s, then. According
to the general rule describing the sound
change, earlier Latin s became r intervocalically.
This particular Latin sound
change is called rhotacism.
Other evidence for Latin rhotacism,
aside from irregular paradigms, can be
found by comparing Latin words which
show rhotacism to their cognates in other
Indo-European languages, which do not.
These cognates show an s where Latin has
r. For example, the stem of the verb “to
be” in Greek is es-, as we can see from the
third singular present form esti. The same
is true in Sanskrit, where the corresponding
form is asti. We even have historical
evidence that indicates approximately
when rhotacism took place: Cicero tells us
that Papīrius Crassus, dictator in 339 BCE,
was the first of his family to stop spelling
his name “Papīsius.”
I would like to leave you with some food
for thought, dear reader. First, what other
verb and noun paradigms can you think of
that show the effects of rhotacism? Second,
what other examples of sound changes in
Latin and Greek grammar can you find?
To begin with, I can tell you that a different
sound change affected intervocalic s in
Greek. What was this sound change? As
with Latin, you might start by looking at
the verb “to be.” Good luck!
About the Author
Christina Skelton is a second-year graduate
student in the Indo-European Studies
program at the University of California,
Los Angeles. She works on the historical
linguistics of Latin and Greek.