:::: MENU ::::

The Garigliano Bowl and Paradigm of Latin Sum ‘I am’


The Garigliano Bowl and Paradigm of Latin Sum ‘I am’

Picture of the Garigliano Bowl


by Todd Clary, Ph.D.

Hello and welcome to the Carmenta Online Latin School blog. This post presents some illuminating information about the common Latin verb sum ‘I am’ brought to light on an Archaic Latin inscription on the Garigliano Bowl. The circumstances of the discovery of this bowl are not certain–-it was in private hands for several years before scholars at the University of Naples brought it to the attention of M. Cristofani. It is said to have been discovered during excavation of a sanctuary of the Goddess Marica near the mouth of the Garigliano River. Excavations were taking place there starting in 1926. The bowl is currently in Naples at the Museo Archeologico Nazionale. Cristofani dated the bowl and inscription to the first half of the 5th cent. BCE (500-450 BCE), though some modern scholars think it might be a little bit older. Here are a few pictures of the bowl itself.

Archaic Latin inscription on the Garigliano Bowl
Exterior with the name of the owners/dedicators
of the bowl the AHUIDIES = Latin Audii/Audeii/Avidi.


Internal view of the Garigliano Bowl
Internal view of the bowl with EZOM ‘I am’ the predecessor
to Latin sum


Drawing of the Interior of the Garigliano Bowl with rendering of the inscription
Drawing of the Interior of the bowl with rendering
of the inscription (on the left one can read MEDEZOMKOM)


Here is a transcription of the inscription around the interior lip, along with a rendering in Classical Latin and translation into English:

Transcription with word breaks:

Classical Latin version:
nē pare (nōlī capere) mē! sum cum meīs sociīs tribus Audiōrum duōrum

English translation:
Do not take me! I am with my three companions (property) of the two Audii

This inscription has many interesting forms and if the reader would like to read a fuller treatment of the whole thing they may consult, for starters, Brent Vine’s 1998 article in the Zeitschrift für Papyrologie und Epigraphik 121:257-62. The form that concerns us in this short post is the form ESOM, which looks to mean ‘I am’, and must be an early form of Classical Latin sum.
Now, in order to put this form in context, let’s take a look at the whole paradigm for the present indicative active of this verb:

Singular: Plural:
1st person: sum 1st person: sumus
2nd person: es 2nd person: estis
3d person: est 3d person: sunt

Probably the most striking peculiarity of this paradigm is that some of the forms start with #es-, but others start simply #s-. This naturally brings up the question: where does the initial #e– belong, and where does it not belong? A quick look at the Sanskrit paradigm of the same verb, or more precisely cognate verb, will help us answer this question:

Singular: Plural:
1st person: asmi 1st person: smas
2nd person: asi 2nd person: stha
3d person: asti 3d person: santi

This paradigm clearly shows that a vowel belongs before the ‘s’ of the root in the singular, but not in the plural. When we remember that Sanskrit merged the vowels ‘a’, ‘e’, and ‘o’ into ‘a’ it becomes clear that the Latin forms es ‘you are’ and est ‘(s)he is’ preserve an ‘e’ inherited from Latin and Sanskrit’s mother tongue, Proto-Indo-European (PIE).

So, given that information, we can reconstruct a Proto-Latin paradigm of sum as follows:

Singular: Plural:
1st person: esmi 1st person: sme/os
2nd person: esi 2nd person: ste
3d person: esti 3d person: senti

Comparing these forms with the classical forms one will doubtless notice many differences that I cannot discuss in full here, but I do need to discuss the loss, or apocope of final short –i# in Latin. This apocope appears to have occurred relatively late (in terms of prehistoric Latin), since an archaic form in the Carmen Saliare (chant of the jumping priests), TREMONTI ‘they are trembling’ apparently preserves it. The loss of final ‘-i# in esmi would have resulted in a vocalization of newly final ‘m’ in esm, which then would have become –em# or –om#. Hence, the form ESOM on the bowl looks to be an archaic preservation. This supposition is reinforced by a phrase in an Old Campanian inscription reading BRUTIES ESUM ‘I am Brutius’. Old Campanian, however, is even more intriguing when comparing ESOM with sum because it attests both ESUM and SUM. This Old Campanian variation makes it tempting to postulate that at some point in the history of Latin there was similarly a coexistence of two forms for ‘I am’, esom and *som. Note that the raising of short ‘o’ to ‘u’ in final syllables is an extremely well documented change so that one could also look on this as being a variation between esum and sum. In fact, the Roman scholar Varro was clearly aware of both variants since he cites the form esum at de Lingua Latina 9.10

These two variant forms for the same word could be reconciled by supposing that at sometime in the history of Latin there was a deletion, or aphaeresis of initial #e-, which, taking some time to diffuse, resulted in variation for a while before one form took over completely. However, such a supposition is rendered impossible by the other forms with initial #e– (es, est) in the very paradigm of sum, as well as countless other forms in Latin with an initial #e-. So how do we explain the supposed coexistence esum and sum and the eventual elimination of esum in favor of sum?

I believe that the following passages from Plautus suggest a solution that involves two forces working together: prodelision and intra-paradigmatic analogy.

nequam esse oportet cui tu integumentum inprobu’s (Bacch. 602)
He must be worthless, for whom you are the shameless bodyguard.

quod quisque in animo habet aut habiturust sciunt (Trin. 206)
They know what everyone has or will have in mind.

These passages suggest that in spoken Latin at the time of Plautus next to the forms es ‘you are’ and est ‘(s)he is’ there were also contracted clitics ’s and ’st. So we could set up a hypothetical Latin paradigm in which clitics coexisted with the full forms:

Singular: Plural:
1st person: esum/’sum 1st person: sumus
2nd person: es/’s 2nd person: estis/’stis?
3d person: est/’st 3d person: sunt

We would then have to explain why the 1st person clitic ’sum was generalized, eventually becoming the only form, while the 2nd and 3rd person clitics ’s and ’st were not. This is when most scholars invoke analogy, noting that sum as the dominant 1st person singular would have at every stage been buttressed by sumus as the only 1st person plural. Thus, analogy with sumus would have given sum that extra push it needed to eliminate esum altogether. For some, this has been the end of the story, but the keen observer might notice that a similar analogy could be set up for the est/’st variants. After all, the 3rd person plural sunt would also never have had an initial #e-. It then becomes necessary to explain why 3rd person clitic ’st, and for that matter 2nd person ’s, which actually appears to have exerted analogical influence on the plural estis, rather than the other way around, never became generalized as the only forms. To solve this problem one could reference the cross-linguistic structure of words; while sum has a vowel in it and adheres to a very common word structure of consonant-vowel-consonant, it is not nearly as common in human languages that we find full-fledged lexemes of the structure consonant-consonant, as *st would be, much less lexemes composed of a single consonant, as *s would be.

So it would appear that in older Latin esom was the full form of ‘I am’, but that it coexisted with its clitic form, first ’smi, before apocope of final –i#, then ’som. Sometime before the classical period sum came to be perceived as the only form of the word and so it has passed down to us in all the classical texts.


todd clary
Dr. Todd Clary, a specialist in the Latin, Greek, and Sanskrit languages, holds a Ph.D. in Classics and Indo-European Linguistics from Cornell University, where he now works as a professor in the Classics Department. He has taught “Homeric Philology”, “Latin Comparative Grammar”, “Archaic Latin and the Italic Dialects”, as well as all levels of Greek, Latin, and Sanskrit. Dr. Clary currently teaches the “Latin IV” and “Latin VI” classes and tutors Latin, Greek, and Sanskrit for the Carmenta Online Latin School.

Click here to see Dr. Clary’s full profile on the Carmenta Faculty Page.



One Comment

  1. Aleksandr Dziuba

    Hello! I am a Lecturer of Romance Philology in Russian University. Your page is very useful, I love it. Thank you very much! Best Regards, Alex