I have little to no doubt that looking up phoneme inventories of various languages, i.e. the sets of meaningful speech sounds they use, might comfortably be one of top-ten activities to procrastinate with at 3 a.m. for many readers of this magazine. Unless you’re more into syntax and morphology when it comes to your night-time fun, which is good, in fact: today’s story might actually be new to you.
Anyway, if you look at the phoneme inventory of Spanish, and specifically its consonants, you might notice it has got fewer of those than its Western Romance siblings: where Spanish makes do with 17–19, others have around 20 (Portuguese, French), or even some 23 (Italian, Catalan). Which makes you wonder if it’s because of a few small reasons that coincided, or if there has been a major tiny phonetic catastrophe that left a bunch of Spanish consonants closing the lids on their coffins so hard that the priest’s Latin went Vulgar: remember that all of the languages I’ve mentioned are close relatives, and were the same language not too long ago, there must have been something going on in Spanish!
Yes, there was! And yes to both the hypotheses: a few notable smaller changes in the consonant sounds took place as Mediaeval Spanish transitioned to Modern Spanish, with one tough chunky big-shot change, which we’ll talk about today. Another look at the consonant phonemes of Spanish reveals that there is no /z/ (as in English ‘zed’), or /ʤ/ (English ‘job’), or /ʃ/ (‘shoot’), or /ʒ/ (last sound in ‘rouge’). All of these, and more, exist pretty much everywhere else in Romance languages, and all of these belong to a group of consonants called the sibilants. Sibilants (Latin for ‘whistling’) include all affricate and fricative sounds1 that are produced by the front of the tongue contacting or coming close to an area ranging from the upper teeth to the hard palate.
Today’s Spanish has only two or three of those: /s/ in ‘sapo’ (equivalent to the English /s/ in ‘smug’), /tʃ/ in ‘chico’ /tʃiko/ (similar to English /tʃ/ in ‘chap’), and sometimes /θ/ as in ‘cero’ /θero/ (like the first sound in ‘through’). There were seven in the Middle Ages. What happened? 1500s did. One of those seven is actually a product of earlier developments, and was not affected by the changes we’re looking at now: it’s /tʃ/, so let me conveniently forget about it and treat you to a table of the rest of the sibilants that Mediaeval Spanish had to offer2:
| /ts/ written as ‘ç, c’ Voiceless dental affricate | /s/ written as ‘ss’ Voiceless alveolar fricative | /ʃ/ written as ‘x’ Voiceless pre-palatal fricative |
| /dz/ written as ‘z’ Voiced dental affricate | /z/ written ‘-s-’ Voiced alveolar fricative | /ʒ/ written as ‘j, g’ Voiced pre-palatal fricative |
If you have any practical knowledge of Spanish, you might notice that the letters ‘c’ and ‘z’ are often pronounced in exactly the same way. Plus, you never see that double ‘ss’ at all, it’s always ‘s’, and always standing for a voiceless sound. If you’re a bit more advanced, you’ll know that there are still words in Spanish that use ‘x’ for a sound that would normally be represented by ‘g’ or ‘j’, the best example of this being ‘México’ pronounced /’mexico/3. Standard spelling calls for ‘Mégico’, but it is what it is, a remnant of how the language used to be.
Anyway, if we take this observation a step further, we can see that it’s the spellings of the voiced sounds collapsing together with those of the voiceless sounds. And that’s one of the things that happened: the voiced merged into the voiceless, /s/ and /z/ became /s/, /ʃ/ and /ʒ/ became /ʃ/. Because this change came from the north of Spain, some scholars attribute it to the influence of Basque, because that language has no voiced sibilants either.
However, before this entire devoicing circus happened, the Affricate Brothers, /ts/ and /dz/, had a disappearing act of their own. In which they disappeared. Leaving no trace but two dental fricatives, voiceless /s̪/ and voiced /z̪/: it is thought to have happened because affricates took a lot of energy to pronounce, and so they gradually ‘weakened’ to the more economical fricatives. But think what you want; I know the Affricate Brothers did it for the art!
Whatever the case, by the beginning of 1600s, this is what Spanish sibilants looked like:
| /s̪/ written as ‘ç, c, z’ Voiceless dental fricative | /s/ written as ‘ss, -s-’ Voiceless alveolar fricative | /ʃ/ written as ‘x, j, g’ Voiceless pre-palatal fricative |
You don’t have to be an eagle-eyed police detective to spot the problem: there were many letters that represented the same sounds at that point, didn’t this result in a horror of homophones? Sure it did: osso (‘bear’) and oso (‘I dare’) were pronounced the same now, and so did coxo (‘limping’) and cojo (‘I take’). These words became homophones and homographs in Modern Spanish, with oso and cojo.
Bad enough? If it is, tough, because the most egregious bit is coming. All of the remaining sibilants were quite similar both acoustically, and in their articulation. See for yourself: they’re all voiceless and fricatives now, and the active articulatory organ is still the same: the tip (or the blade) of the tongue. The only difference is where it lands: at the back of the upper teeth or right behind them (dental), on the rough area behind the upper teeth (alveolar), or somewhere right behind that rough area (pre-palatal). It’s literally millimetres that separate them, and if you’re talking too fast, or are intoxicated (with all the phonetic change, no doubt), or if you’re not very articulate in general, your tongue will slip and result in tragic accidental two- and three-way homonyms. Have a look, all of the Spanish words in the table sounded similar at that point.
| Mediaeval Spanish | English (all nouns) |
| caça /katsa/ -> /kas̪a/ | hunt |
| casa /kaza/ -> /kasa/ | house |
| caxa /kaʃa/ | box |
A possible solution to this is making these sounds more different, and that’s a change that started in the north of Spain practically as soon as all the affricates and voiced sibilants were lost, spread around much of the country throughout the 1600s, resulting in sounds that are part of standard pronunciation in the Spanish of Spain. What they did was take the dental /s̪/ and shift it forward, away from the alveolar /s/, and as far as they could short of sticking their tongues out, getting the voiceless interdental fricative /θ/. Then, it was /ʃ/’s turn to do the same thing, except in the opposite direction: it was now articulated with the back of the tongue, deep in the mouth cavity, with the voiceless velar fricative /x/. Now all three were wonderfully wide apart, both in how they were pronounced and in how they sounded.
| Mediaeval Spanish | Modern Spanish |
| caça /katsa/ -> /kas̪a/ | caza /kaθa/ |
| casa /kaza/ -> /kasa/ | casa /kasa/ |
| caxa /kaʃa/ | caja /kaxa/ |
So the story of a Spanish king lisping and single-handedly causing phonetic change is all rubbish. The whole job was done by the average Joe (or rather José) doing the sensible thing.
But wait, there’s more to the story than we might have thought! What about the Spanish-speaking Americas, what about the south of Spain, or the Canary Islands? They sure as heck don’t have /θ/ in their speech, and most are doing fine without /x/, too!
Let’s set the stage first by noting that the south of Spain was linguistically quite different from the north. Because the south was where the Reconquista ended, these areas experienced more influence of Arabic, and more widespread use of Mozarabic, a set of Romance dialects spoken in the areas of the Iberian Peninsula controlled by Muslims. Distinct from Mediaeval Spanish, they became a substrate for it, as Castile expanded southwards.
This meant that the south had a slightly different story when it came to sibilant changes, with a few important differences from the north, even though the initial set of sibilants was probably the same. First, while the loss of /ts/ and /dz/ affricates did occur, the devoicing of the voiced sibilants was a much later event than in the north. Second, instead of diverging into more distinct sounds, /s̪/ and /s/ converged, resulting in /s̪/ (there were still voiced sibilants present, so /z̪/ and /z/ turned into just /z̪/). Some scholars believe, actually, that due to the influence of the local Arabic, there were no alveolar /s/ or /z/ to begin with, and their dental counterparts had always been there (/ts/ and /dz/ turned into dental fricatives, exactly like up north).
Third, at some point after the devoicing and loss of affricates, /s̪/ and /ʃ/ were the only sibilants in use in this region, and while, acoustically, they are reasonably different, /ʃ/ experienced a shift towards the back of the mouth; interestingly, it was more extreme than in the language of the north of the country, and the resulting sound was the voiceless glottal fricative /h/, articulated far deeper than /x/, at the vocal cords. This is probably due to the fact that /h/, initially present in Mediaeval Spanish, persisted for much longer in the south (up to modern times, in some areas), so the shifting /ʃ/ merged into it, because the historical /h/ was still in use, and was a familiar sound. In the end, only /s̪/ and /h/ were left of the rich sibilant beginnings:
| Mediaeval Spanish | Modern Spanish | Modern Spanish (south of Spain and Latin America) |
| caça /katsa/ -> /kas̪a/ | caza /kaθa/ | caza /kas̪a/ |
| casa /kaza/ -> /kasa/ | casa /kasa/ | casa /kas̪a/ |
| caxa /kaʃa/ | caja /kaxa/ | caja /kaha/ |
While this approach produced many homophones, it was the accent of the Spanish south that spread into the Americas, becoming the way most Spanish-speaking people speak. So, having started with Middle Age fricatives and affricates nobody even uses any more in the language, we have essentially traced the history of one of the most prominent dialectal features of Spanish: seseo, i.e. saying /s̪/ where speakers from much of Spain and, traditionally, Madrid would say /θ/. For a language, 300–400 years isn’t much, and we’re still hearing the echoes of the great phonetic changes that took place during the Spanish Golden Age.
1 A fricative is a consonant sound, in which the organs of speech (lips, teeth, tongue, hard palate etc.) create a constriction in the airflow without obstructions, resulting in a hissing noise (check the first sound of words like ‘see’, ‘far’, ‘shoe’, ‘zoom’, ‘there’). An affricate is a sound, in which the organs of speech do create an obstruction at first, i.e. the air can’t go through, and then release it relatively slowly, to produce some hissing noise like that of a fricative (look for the ‘ch’ sound in words like ‘chin’ or ‘chuckle’, if it sounds something like ‘t-sh’ to your ear, you’ve got the idea).
2 Note that there are variations to this terminology, and I provide the terms that I typically use and that my sources use. ‘Dental’ is often referred to as ‘Denti-alveolar’ (the place of articulation is the back of the upper teeth and the area immediately behind them), ‘Pre-palatal’ is effectively the same as ‘Post-alveolar’ (the place of articulation is immediately behind the alveolar ridge, the rough area behind your upper teeth), ‘Interdental’ sounds are often called ‘Dental’ (the place of articulation is between the teeth, but close to or touching the upper teeth). Plus, for ‘Alveolar’ sounds, the place of articulation is the alveolar ridge.
3 /x/ is the voiceless velar fricative, whereby the back of the tongue comes close to the soft palate at the back of the mouth (the tongue position is similar to that of ‘oo’ in ‘shoo’).
Get more interesting language and linguistics content in our magazine. You can subscribe below
