← Back to blog
5 min readTabla

Espanish and estrong: the ghost E before S words

You say eschool, estrong and espeak without noticing. It is not carelessness, it is a Spanish rule. Here is how to switch it off in three minutes.

  • pronunciation
  • common mistakes
  • accent
Espanish and estrong: the ghost E before S words

You go to say I speak English and out comes I espeak English. Nobody corrects you, because you're perfectly understandable, and that's exactly why you've carried it for years. That ghost vowel in front of the s is the most recognizable marker of a Spanish accent in English.

The cause isn't carelessness or a low level. It's a Spanish rule running on its own: Spanish doesn't allow a word to begin with s plus a consonant, so your mouth slips in an e to repair what it reads as an impossible syllable. The fix is just as mechanical. Start the word on the hiss, with air already moving and your vocal cords still, instead of starting it on a vowel. One motor gesture, trained in minutes rather than months.

Your mouth isn't failing: it's running a thousand-year-old rule

Spanish has been inserting that e systematically for a millennium. Every Latin word that began with s plus a consonant picked one up on the way in: schola became escuela, spiritus became espíritu, sponsa became esposa, a sound change linguists call prothesis. Italian never made it, which is why it says scrivere and spirito without trouble.

English simply hands that same rule new material to work on. As Esther de Leeuw and colleagues put it in Second Language Research (2021): "Spanish native speakers are known to pronounce onset /sC/ clusters in English with a prothetic vowel, as in esport for sport, due to their native language phonotactic constraints."

Diagram of a wall blocking two tightly packed blocks, with a sphere wedged in front of them as a makeshift ramp

The rule fires without you ever opening your mouth. Max Freeman, Henrike Blumenfeld, and Viorica Marian tracked Spanish–English bilinguals processing non-words like esteriors in Frontiers in Psychology (2016): responses landed at a mean of 848 milliseconds when the item followed a cognate prime such as stable. Their reading of it: "Spanish–English bilinguals may activate a phonological epenthesis constraint in the non-target language during comprehension." Your brain inserts the e even while it's listening.

Start on the hiss, not on the vowel

The fix is one physical gesture: stretch the s before you release the rest. Put your tongue in s position, let air run for two full seconds, ssssss, and only then add the consonant behind it: sssss-pain, sssss-trong, sssss-tudent. While voiceless air is hissing out, a vowel physically cannot appear, because vowels need your vocal cords to vibrate.

Two waveforms compared: the top one starts abruptly out of silence, the bottom one carries a soft extra bump added before its onset

Four words give you away daily:

  • school, not eschool. Open on the air; the k arrives half a beat later.
  • Spanish, not Espanish. The word you say most often about yourself is the one that outs you first.
  • strong, not estrong. Three consonants back to back, nothing in front of them.
  • student, not estudent. Spanish estudiante sits right behind it and pushes.

Nobody will ever correct you, and that's the problem. You're in the passport line in Miami, the officer asks what you do, you say I'm a estudent, he nods and waves you through. The mistake costs you nothing in the moment, so it fossilizes.

Hearing the difference does nothing if you can't produce it

Telling spi from espi by ear and producing spi cleanly are two separate skills, and training one doesn't hand you the other. The de Leeuw study found that bilinguals with more English exposure produced the contrast more accurately while their perception stayed flat, with no significant correlation between the two measures. Accurate perception is neither necessary nor sufficient for accurate production.

Listening to more podcasts will not move your tongue. Input trains your ear; the mouth needs its own reps. The meta-analysis by Junkyu Lee, Juhyun Jang, and Luke Plonsky pooled 86 studies of explicit pronunciation instruction and found a large effect, d = 0.89, with the biggest gains where training included immediate feedback.

Could you say Spanish students speak Spanish five times fast right now without a single e slipping in? That's the gesture Tabla drills: you say it out loud against a clock, and the app scores what actually came out of your mouth.

The bridge trick: lean on the word before

Context changes the difficulty, and you can use that. Robert Carlisle documented in Applied Linguistics (1991) that epenthesis shows up far more often when the previous word ends in a consonant than when it ends in a vowel. Big school is a minefield. The school is easy ground, because you arrive with your mouth already open on a vowel.

Two blocks joined by a continuous bridge while a leftover filler piece floats away and fades out

Practice along that gradient. Start with the easy frames, the school, a student, to Spain, and move to the hard ones only once those come out clean: big school, brilliant student, back to Spain. You build the gesture where your mouth pushes back least, then carry it across.

Does it actually matter, or is it cosmetic?

It matters for your accent, not for being understood. Adding a vowel almost never breaks communication: Jennifer Jenkins's Lingua Franca Core, the inventory of features that genuinely carry intelligibility between non-native speakers, explicitly tolerates vowel epenthesis. What it refuses is the opposite: "Omission of consonants in clusters is allowed word-medially and -finally (but not -initially)." Swallowing the s in strain and turning it into train is the expensive mistake.

The reasonable goal isn't sounding native. Tracey Derwing and Murray Munro framed it this way in Language Teaching: "The recent revival of interest in pronunciation research has brought a change of focus away from native-like models toward easy intelligibility." The ghost e is one of the few accent features that a single repeatable gesture actually fixes. That's the reason to go after it: low cost, fast payoff in how you sound.

Say this out loud: "My sister studies Spanish at a small school in Scotland."

Five traps in a row: studies, Spanish, small, school, Scotland. Only a small is easy ground, because you arrive from a vowel. The other four all land after a consonant (sister studies, studies Spanish, small school, in Scotland), which is exactly where the e slips in. Run the sentence again and stretch the hiss on each of those four.

Next time you're about to say school, pause for half a second and start on the air. If the sound opens with a hiss instead of your voice, the e has no way in.

Try this today: record yourself saying "Spanish students study strange subjects" three times in a row and listen only to the start of each word. If you hear an e, stretch the s for two seconds and go again.