Stellar Stars


Here’s an interesting question that I was sent to me by email:
I am curious as to why some of the languages that developed from Latin had to put an extra ‘e’ at the start of some of their words.

Here are some examples:

Latin Italian French Spanish English
stēlla stella étoile estrella star
status stato état estado state
spero speranza espère esperanza hope
spōnsa sposa épouse esposa wife

It looks as if the Gauls, and the people living in the Iberian peninsula, couldn’t cope with the st- and sp- beginnings, and had to stick an ‘e’ on the front. Is this because words in the Celtic languages they spoke didn’t have such beginnings? I can’t find any similar words in modern Welsh.

Incidentally, the words for hope have a cognate in English – esperance, which is a old word for hope or expectation [source], and the ones for wife have a cognate in spouse (husband, wife).

Let’s look at the origins of some of these words to see how they have changed over time.

The Latin word stēlla (star), comes from the Proto-Italic *stērolā (star), from Proto-Indo-European *h₂stḗr (star). This became estoile/esteile/estelle in Old French, and estoile in Middle French. It was (e)strela in Old Portuguese and estrella in Old Spanish So the extra e has been there for a while [source].

In Proto-Celtic the word for star was *sterā, from the same PIE root as the Latin stēlla. This became *ster in Proto-Brythonic, Old Breton and Old Cornish, and ster in modern Breton and Cornish. So at least some speakers of Celtic languages could cope with the initial st-. In Old Welsh it was *ser, in Middle Welsh it was ser / syr, and in modern Welsh it’s sêr. It was also borrowed into Old Irish as ser [source].

The Latin word status means state, status, condition, position, place or rank. It became estat in Old French, from which we get the English word estate. Meanwhile in Old Spanish it was (e)strela, and in Old Portuguese it was estado [source].

It was borrowed into Old Irish as stad (stop, stay, delay), which is the same in modern Irish [source]. Proto-Brythonic borrowed it as *ɨstad from the Vulgar Latin *istatus, this became (y)stad / (y)stât in Middle Welsh and ystad (state, condition, situation) in modern Welsh [source].

Do any of you have any thoughts on this question?

5 thoughts on “Stellar Stars

  1. The phenomenon is called prothesis: in order to resolve a “consonant-heavy” cluster like str, a vowel is added at the beginning, which allows the cluster to be spread across two “lighter” syllables (e.g. stra => es-tra). The process is prominent in several Romance languages (Spanish perhaps being the most well-known), but occurs in unrelated languages too, such as Syriac Aramaic, where the traditional alphabet is called Estrangela, a loan From Greek strongule “rounded” with a prothetic e added at the beginning.
    Why some languages have the need to resolve such clusters and others do not, I wouldn’t know. I guess that’s just to do with the overall phonetic rules of the language at hand.

  2. Sometimes languages develop phonotactic restrictions and when words from other languages without the same restrictions are incorporated, these loanwords undergo certain changes so as to fit into the phonotactics of the receiving language. Prothesis is one strategy (in Hebrew: estratégya) for that. In Polynesian languages, consonant clusters are broken up by means of intervening vowels, e.g. cream, in Tuvaluan: kulimi. Another way to deal with impermissible consonant clusters is deletion (or elision). /kn/ is such a cluster in English: knee (German: Knie) is pronounced [ni:] so that orthograpically the is still there but is not anymore in the spoken language.

  3. There are some examples in Welsh where words of Latin origin beginning with a consonant clusters have an initial y added: ysgol (>scola), ystafell (>stabulum), ysgubor (>scopa), ysbryd (>spiritus), ysgrifennu (>scribendum).

    According to , there are also words in Welsh beginning ysb, ysg and yst with their origins in Proto-Celtic, in which they did not have an initial vowel (e.g. ysgafn > P.C. skamos – but I have no idea what that supposition is based upon. I would hazard a guess that the ‘need’ for an initial vowel before such consonant clusters was variable among Celtic languages, as it is among Romance (and probably other) languages, rather than being a specifically Celtic feature. In colloquial Welsh, the initial y is often dropped in some of the above examples – e.g. sgrifennu (or sgwennu). It is always retained in a stressed syllable (e.g. ysbryd) but can be dropped where the addition of a suffix shifts the stress to another syllable, hence sbrydion.

  4. I should add that (according, again, to , some of the Latin-derived words had already acquired their initial vowel in their Vulgar Latin forms – which could have been a Celtic legacy, or maybe even a regional feature pre-dating Celtic.

    It is also worth noting that initial consonant clusters of this type are dealt with differently in other languages. For example, in German, initial sp and st are pronounced /ʃp/ and /ʃt/, respectively – as they are in certain Irish regional varieties of English and (so my father told me) in some varieties of Italian. (German orthography lacks initial sn altogether, with German cognates of English words with initial sn being rendered with schn – and pronounced /ʃn/).

Leave a Reply

Your email address will not be published. Required fields are marked *