Seville Spanish Accent: The S Survives 2.8% of the Time

By Camila Rossi, Culture writer, Buenos Aires & Barcelona

Updated September 4, 2026 · Reviewed by the Audaz editorial team

Every travel guide says the same thing about Seville. They drop their S and it sounds charming.

That is not wrong. It is just the least useful true sentence you can say about an accent you are about to stand in the middle of. It tells you nothing about when the S goes, what replaces it, or why the woman at the ticket desk sounds different from the man selling you churros.

Linguists have measured this city. Three studies built on the same corpus give us tens of thousands of individual sounds, counted by hand. We read them so we could answer the only question that matters to a learner: what will you actually hear?

What did the linguists actually count in Seville?

The evidence here comes from PRESEEA, a long-running project that records ordinary people talking in Spanish-speaking cities and transcribes what they say. The Seville arm is called PRESEEA-Sevilla.

Three published studies sit behind this article. Here is what each one measured.

These are not vibes from a language blog. They are hand-coded counts, published in peer-reviewed journals, with the statistics shown. That is why we can give you percentages instead of adjectives.

Does Seville really drop every S?

Close enough to be worth saying, and far enough off to be worth correcting. In the full 72-interview study, the syllable-final S came out as a clean, full S only 2.8% of the time.

But it does not vanish into nothing. It splits two ways.

Those first two lines are the whole problem. Half the time there is a sound to catch. Half the time there is not. If you were taught to listen for the breathy H, you were taught half an accent.

Why does the S disappear in one word and survive in the next?

Because position decides. An S in the middle of a word behaves nothing like an S at the end of one.

The 2023 study counted 4,176 sounds among university-educated speakers and split them by position. Word-internal cases, like the S in hasta, numbered 968. Word-final cases, like the S in gracias, numbered 3,208.

Read those two rows again. Inside a word the S almost always leaves a trace you can hear. At the end of a word it usually does not.

The study puts a sharper number on it. Of every S that disappeared completely, 99.6% disappeared at the end of a word. Word endings are where Spanish keeps its grammar. That is the part Seville deletes.

Where is the hardest place to hear a Sevillian S?

At the end of a sentence, right before the speaker stops. This is the finding that should change how you practise, and almost nobody mentions it.

The 2023 study sorted every word-final S by what came next: a vowel, a consonant, or a pause. The pause column is brutal.

Before a pause, the breathy H you were told to listen for happens 4.8% of the time. A full, clear S is nearly twice as likely at 8.7%. The standard advice is backwards in exactly the spot where a learner is concentrating hardest.

The other two contexts behave differently again.

There is a practical rule buried in that table. Two thirds of the S sounds you meet come before a consonant, and that is the one context where the breath is the likeliest thing to happen. Train your ear on the middle of sentences and you will do well. Train it on the ends and you will hear silence.

Does the plural S survive because it carries meaning?

This is the assumption almost every learner makes, and the corpus says no.

The idea has a name in linguistics. It is called functional load. Some S sounds carry information. The S in los libros marks a plural. The S in hablas marks you. The theory says speakers should hold onto those harder than an S that carries nothing.

The 2023 team tested it. They coded whether each S was a grammatical marker, ran it through the model, and found no significant effect. They dropped the variable from the final analysis. They also note that a study in Granada reached the same conclusion, and that researchers working on Tenerife left the variable out for the same reason.

Seville deletes the plural S as readily as any other S. Your plurals are not protected. They are carried by the article and by context, not by the ending of the noun.

That has a direct consequence for you. La casa and las casas can land on your ear as the same string of sounds. The word doing the work is la or las, and in fast speech even that S is at risk. Listen for the vowel and the article, and let the noun ending go.

Does it matter who you are talking to?

Yes, and education is the clearest predictor of the three that were tested.

The 72-interview study split speakers by how much formal education they had. The two halves of the S trade places as you move up.

Nobody in that table is producing textbook S sounds. The difference is whether the sound is softened or deleted. The more schooling a speaker has, the more likely you are to get something audible.

Age turned out to matter far less. Among educated speakers, all three generations aspirated at around 49%. The one small drift is in the full S, which rose from 2.5% in the youngest group to 5.6% in the middle group and 6.7% in the oldest. The authors read that steadiness as a sign the pattern is settled rather than moving.

Sex showed a small, consistent split among educated speakers. Men preferred the breath, at 54.8%. Women preferred deletion, at 47.6%, with the breath close behind at 43.4%.

Is Seville still a seseo city?

Less than it was, and this is the part the guidebooks have not caught up with.

A quick definition first. In most of Spain, casa and caza sound different, because Z and soft C are pronounced like the TH in think. That is called distinción. In seseo, both are pronounced as S, so cerveza comes out closer to serve-sa. Seville has been the historic capital of seseo in Spain.

The 2016 study counted 3,246 relevant sounds from 24 university-educated speakers recorded between 2009 and 2015. Seseo accounted for 25.54% of them. The northern pronunciation accounted for 74.46%.

The author compares that with earlier work on the same city. Studies of 1970s recordings put seseo at 70% and 74%.

Nearly four in five educated Sevillians switched between both pronunciations while talking to the same interviewer, about the same everyday topics. There is no single Seville accent to learn. There is a range that one person moves through inside one conversation.

What can these numbers not tell you?

Six things, and they matter as much as the findings.

How do you train your ear for Seville before you go?

Four drills, each aimed at something the data says will actually happen to you.

Want the wider map first? Our region-by-region guide to how Spanish sounds across Spain sets Andalusian next to Castilian, Canarian and the rest. The complete guide to Andalusia covers what to do once you can hear it. And our guide to the Feria de Abril covers the week when the city is loudest.

This article is the third in a series that measures a city instead of describing it. We have already done the Buenos Aires accent and why porteños say SH. We have also done Mexico City, where the neighbourhood-accent story turned out to be a myth. For the rules underneath all three, start with our Spanish pronunciation guide. And our Spanish food pronunciation piece fixes fifteen words you will say in a bar.

So what is the honest description of the Seville accent?

It weakens almost every syllable-final S, keeping a clean one under 3% of the time. It usually leaves a breath inside a word and usually leaves nothing at the end of one. It deletes grammar endings without mercy, because grammar does not protect them. It deletes hardest right before a pause.

And its most famous feature, the seseo that made the city the reference point for half the Spanish-speaking world, is now the minority pronunciation among its own graduates.

The charm is real. It is just not where the guidebooks put it. It is in a city that runs three pronunciations at once and lets each speaker choose, sentence by sentence.

Frequently asked questions

Sources

• León-Castro Gómez, M. (2025). La variación de la /s/ posnuclear: patrones sociolingüísticos en el español de Sevilla. Nueva Revista del Pacífico, 83, 355-382. Read the study.
• Jiménez Fernández, R. & León-Castro Gómez, M. (2023). Variación en la pronunciación de la /s/ implosiva: datos del nivel sociocultural alto en el corpus PRESEEA-Sevilla. Boletín de Filología, LVIII(2), 371-398.
• Santana Marrero, J. (2016). Seseo, ceceo y distinción en el sociolecto alto de la ciudad de Sevilla: nuevos datos a partir de los materiales de PRESEEA. Boletín de Filología, LI(2), 255-280. Read the study.