Romance: an example of a language sub-family with common descent

Back to the Syllabus

Let's start today simply with a list of words in several Western European languages.

English

Italian

French

Spanish

Portuguese

rain

pioggia

pluie

lluvia

chuva

lead (metal)

piombo

plomb

plomo

chumbo

wound (from a blow or cut)

piaga

plaie

llaga

chaga

full

pieno

plein

lleno

cheio

weep

piangere

pleurer

llorar

chorar

call

chiamare

clamer

llamar

chamar

key

chiave

clef

llave

chave

 Note: Italian "pi" is pronounced "py"; Italian "chi" is "ky"; Spanish "ll" is a postalveolar fricative, sometimes close to a fricative English consonatal "y"; Portuguese "ch" is "sh." Not all these words are the common equivalents: "wound" is more typically "herida" in Spanish and "ferida" in Portuguese, and the French equivalent of "llamar" is "appeler" ("clamer" means "to cry out"); but all these words are fairly basic in these languages.

 

 

The first thing to notice is how different English is from these other languages. English obviously belongs to a different language group, and if one had only these eight words to go on, one might not suspect that English had any familial relation to the others. In each case, the common English word is a Germanic word not closely related (if at all) to the Romance word. OTOH, we have many borrowings from these Romance words: "pluvious" for rainy; "plumb" and "plumber," "plenitude," "claim" and "clamor," --and "clef" won't set off your spell-check, at least in a musical context.

The second thing, more important, to notice is the clear family relation among Italian, French, Spanish, and Portuguese. Such a relation in fact becomes clear only from accumulated impressions. If we had only the word for "rain" we couldn't be very sure. But an extended list establishes certain sound equivalences. Italian "pi" = French "pl" = Spanish "ll" = Portuguese "ch." But not everywhere that Spanish "ll" matches Portuguese "ch" do we have "pi/pl" in the others; we might have instead "chi/cl." Here and there there's something that runs counter to form, like "plomo," Spanish for "lead."

How does a linguist, faced with a family of languages, reconstruct the vanished parent language? We could, tentatively, note that this family of languages falls into two groups: one that has a stop consonant at the start of these words (French and Italian), the other that has a fricative (Spanish and Portuguese). We would know from external evidence that Spanish and Portuguese are spoken in Iberia, and French and Italian to the east; this would strengthen our idea that there are two sub-sub-families at work here.

The fact that the "homeland" of these languages is toward the east--in Italy--wouldn't decide whether the fricative or the stop was the original initial consonant, because the western languages might just as well have been innovative as conservative. To form a theory about this issue, a linguist essentially asks what change is more likely, is "easier" in the broadest sense. Is it more likely that speakers would have taken a simple fricative and made a stop or stop + semivowel out of it (if the Spanish/Portuguese sound were original), or is it more likely that a complex sound would have moved towards different, simpler manifestations in the daughter languages?

I've phrased that in a leading way to get you to say the "right answer": it's much more likely that the original sound was, all else being equal, the most complex of the descended options. In this case, that would be either the Italian /py/ or the French /pl/ sound. Which of these would be original? The Italian sound places the tongue at the front of the mouth; in French, to make the /l/, the tongue is back up against the palate or behind the alveolar ridge--close to where it is in Spanish or Portuguese, to make the fricatives there. So a reasonable reconstruction is that the French version of the initial sound in "pluie," "plomb," &c. is close to the "original" sound.

That's a gross overgeneralization of what reconstructive historical linguists do. In the case of Romance, no linguist has to do this to reconstuct the ultimate original; the parent language is Latin, and its classical equivalent terms for these words are pluvium, plumbum, plaga, plenus, plorare, clamare, and clavis. But much intermediate evidence is missing. There are obviously several steps between Latin and the modern Romance languages, and the data in between are missing; the Romance languages form subfamilies (especially the southwestern subfamily represented by Spanish, Portuguese, and several minority languages), and we do not have good records for the early periods of those languages (since written Latin, as we'll see next time, remained stable while the vernacular spoken subfamilies were diverging).

Also, things are never as simple as they might seem. My simplified table above might lead one to guess that one could generate a modern equivalent in any language for Latin words that begin with "pl" just by running through a standardized set of sound changes. But what's going on with Spanish "plomo" ("lead")? And what about this next set of words:

English

Italian

French

Spanish

Portuguese

planet

pianeta

planète

planeta

planeta

plastic

plastico

plastique

plástico

plástico

Pluto

Plutone

Pluton

Plutón

Plutão

plebeian

plebeo

plébéien

plebeyo

plebeu

 If I'd given you these equivalents first, you might think that English was a very recently diverged cousin from the Romance languages. You might also think that the Romance languages are hardly separate languages at all--at most, slightly divergent dialects or "accents."

The reason for these close resemblances is obvious after some thought: English and these four Romance languages have similar words here because these words and concepts, unlike the basic and concrete words in the first table, are specialized and/or recent. "Planet," "plastic," and "Pluto" have all been borrowed from Greek in relatively recent times. "Plebeian" is Latin, and refers to a specific social rank in ancient Rome; the use of the term disappeared with the decay of the Roman state, and the word was revived only as a learned term in the 1500s to describe a Western European social class in terms of one of the classes of ancient Rome. The word entered these languages with the form "pl" long after the system of sound changes had erased the older Latin "pl" from its daughter languages. In fact, words beginning with "pl" in Italian, Spanish and Portuguese all tend to be relatively new and learned words borrowed from Latin and Greek.

So why do Italians say "pianeta" and Spanish speakers say "plomo"? The first rule of historical linguistics is that there's always an exception. "Planet" entered the Western languages somewhat earlier than "plastic," in the 1300s or earlier as opposed to the 1600s, so it probably entered Italian early enough to be assimilated to words of the form "pioggia." (Note that this fact also offers indirect evidence for the original "pl" sound in "pioggia"; the "pl" may have persisted till very late times in Italian.) "Plomo" is like "planeta": it's a learned term imported into Spanish from Latin at a relatively late date. We'd expect a Spanish word something like *llumo, *llumbo, *llomo, and on the evidence of Portuguese "chumbo," there probably was such a word, superseded by the learned borrowing of the Latin word for "lead," the chemical element "plumbum" (Pb).

Generalizations

Noise can come from cultural importations and overlays (like the planet / plastic words above). Noise can also come from substrate elements. By "substrate," we mean the languages spoken in a given territory before a new language arrived. Take, for example, these very common words in Romance:

English

Latin

Italian

French

Spanish

Portuguese

Romanian

bed

lectus

letto

lit

cama (also lecho)

cama

(but of a river: leito)

pat

dog

canis

cane

chien

perro (also can)

cão

ciine

 

We can establish more of what we already know: that English is way different from the common Romance vocabulary (though again there are English cognates and borrowings: "litter," oddly enough, and also the cognate "hound" and learned borrowing "canine"). The Romance languages descend from Latin. But somehow, despite the commonness of these terms, Spanish and Portuguese have adopted new common words for "bed," and Spanish has a new common word for "dog." Since "cama" and "perro" come neither from Latin nor from any known later influence on the languages, we assume that they come from a now-lost substrate: a language spoken in the territory that Roman settlers moved into, from which a few words entered the conquerors' language before the substrate language disappeared. (Compare the persistence of a few American Indian words in English, a theme we'll return to later this semester.)

Brief Overview

I've been most interested in the Romance languages as a general illustration of linguistic divergence. But they are interesting in themselves, not least because English contains so many Romance elements via French. The Romance subfamily contains 46 recognized varieties, though only French, Spanish, Italian, Portuguese, and Romanian have the status of modern national languages with literary and governmental traditions. Catalan (in Spain) and Provençal (in France) have substantial literary traditions and some recognition as political minority languages; other Romance languages have distinct minority status. Go to Romania Minor for a look at some of these minority varieties of Romance; the opening page of that site has a fascinating parallel-text passage that allows for some comparisons among lesser-known Romance languages.