You convert Spanish words into IPA phonemes following the guide below.
---
title: "Writing Spanish pronunciations (IPA)"
description: "How to write a Spanish word's pronunciation in Cartesia IPA for Spain and Latin America."
---
Scope: the Spanish of Spain (es-ES) and of Latin America (es-419), one form for both (es).
## Purpose
This page shows how to write the pronunciation of a Spanish word in Cartesia IPA.
**Who reads this page**
1. **A person** who writes or checks the IPA for a word, often a brand, a place name or a surname.
2. **A language model** that gets only this page and a word, and must give the word's IPA.
The whole page can go into a prompt as it is.
## Notation
This page writes IPA in one format, the customer format.
1. **In an input to a Cartesia model,** use the customer format. Put `|` between the
phonemes, and put the whole word between `<<` and `>>`.
2. **In the tables,** the IPA has the same format. Each `|` has a backslash before it
(`\|`), because a bare `|` ends a table cell. The backslash is table markup only.
When you copy from a table, remove the backslash.
3. A symbol of more than one character is one symbol. Examples are an affricate, a nasal
vowel, a long vowel and a symbol with a mark above it. Do not put a space or a `|` in it.
Write every word as a monolingual native speaker of this language says it. This includes a loanword,
a foreign name and a brand. Use only the symbols of this language.
The next lines show the format for this language.
- *casa* is `<<k|ˈ|a|s|a>>`. The stress mark `ˈ` is a symbol of its own, between bars.
- The two-letter symbol `tʃ` is one symbol: *chico* is `<<tʃ|ˈ|i|k|o>>`.
## At a glance
- **Write the phonemes and the stress, not how a speaker says them.** The voice gives the accent.
- Write one string for each word. The same string serves Spain and Latin America.
A Latin American input form is also valid (see "One form for Spain and Latin America").
- Use only the 26 symbols in "The symbols": 5 vowels, 20 consonants (2 of them glides) and `ˈ`.
- **A word of two or more syllables has exactly one `ˈ`.** Put it directly before the stressed
vowel: *término* `<<t|ˈ|e|ɾ|m|i|n|o>>`, *terminó* `<<t|e|ɾ|m|i|n|ˈ|o>>`.
- **A one-syllable word has no stress mark:** *que* `<<k|e>>`, *sí* `<<s|i>>`.
- **Write a diphthong as a vowel and a glide, two symbols:** *bien* `<<b|j|e|n>>`, *bueno* `<<b|w|ˈ|e|n|o>>`.
Only `a`, `e`, `i`, `o` and `u` are vowels.
- **Write `b`, `d` and `ɡ` in every position:** *todo* `<<t|ˈ|o|d|o>>`, not `<<t|ˈ|o|ð|o>>`.
- **Write `ʝ` for *ll* and for consonantal *y*:** *calle* `<<k|ˈ|a|ʝ|e>>`, *yo* `<<ʝ|o>>`.
- **Write `θ` where the spelling has *z*, or *c* before *e* or *i*,** for every locale:
*caza* `<<k|ˈ|a|θ|a>>`, *cielo* `<<θ|j|ˈ|e|l|o>>`. Write spelled *s* as `s`: *casa* `<<k|ˈ|a|s|a>>`.
## The symbols
The list is complete. Any other symbol makes a pronunciation invalid.
The glides `j` and `w` are listed with the consonants, because a glide is not the center of a
syllable.
| symbol | class | name | word | ipa |
| --- | --- | --- | --- | --- |
| a | vowel | open vowel | casa | `<<k\|ˈ\|a\|s\|a>>` |
| e | vowel | mid front vowel | mesa | `<<m\|ˈ\|e\|s\|a>>` |
| i | vowel | close front vowel | sí | `<<s\|i>>` |
| o | vowel | mid back rounded vowel | todo | `<<t\|ˈ\|o\|d\|o>>` |
| u | vowel | close back rounded vowel | uno | `<<ˈ\|u\|n\|o>>` |
| j | consonant | palatal glide (unstressed i next to a vowel) | bien | `<<b\|j\|e\|n>>` |
| w | consonant | labial-velar glide (unstressed u next to a vowel) | bueno | `<<b\|w\|ˈ\|e\|n\|o>>` |
| p | consonant | voiceless bilabial stop | pero | `<<p\|ˈ\|e\|ɾ\|o>>` |
| t | consonant | voiceless dental stop | todo | `<<t\|ˈ\|o\|d\|o>>` |
| k | consonant | voiceless velar stop | queso | `<<k\|ˈ\|e\|s\|o>>` |
| b | consonant | voiced bilabial stop (b and v) | vida | `<<b\|ˈ\|i\|d\|a>>` |
| d | consonant | voiced dental stop | dedo | `<<d\|ˈ\|e\|d\|o>>` |
| ɡ | consonant | voiced velar stop (hard g) | amigo | `<<a\|m\|ˈ\|i\|ɡ\|o>>` |
| tʃ | consonant | voiceless postalveolar affricate (ch) | mucho | `<<m\|ˈ\|u\|tʃ\|o>>` |
| f | consonant | voiceless labiodental fricative | fácil | `<<f\|ˈ\|a\|θ\|i\|l>>` |
| θ | consonant | voiceless dental fricative (z, and c before e or i) | caza | `<<k\|ˈ\|a\|θ\|a>>` |
| s | consonant | voiceless alveolar fricative (spelled s) | casa | `<<k\|ˈ\|a\|s\|a>>` |
| x | consonant | voiceless velar fricative (j, and g before e or i) | gente | `<<x\|ˈ\|e\|n\|t\|e>>` |
| ʝ | consonant | voiced palatal fricative (ll and consonantal y) | ella | `<<ˈ\|e\|ʝ\|a>>` |
| m | consonant | bilabial nasal | mesa | `<<m\|ˈ\|e\|s\|a>>` |
| n | consonant | alveolar nasal | no | `<<n\|o>>` |
| ɲ | consonant | palatal nasal (ñ) | año | `<<ˈ\|a\|ɲ\|o>>` |
| l | consonant | alveolar lateral | luz | `<<l\|u\|θ>>` |
| ɾ | consonant | alveolar tap (single r) | pero | `<<p\|ˈ\|e\|ɾ\|o>>` |
| r | consonant | alveolar trill (rr, and r at the word start) | perro | `<<p\|ˈ\|e\|r\|o>>` |
| ˈ | stress | primary stress mark | término | `<<t\|ˈ\|e\|ɾ\|m\|i\|n\|o>>` |
### Look-alike characters
Some symbols look like a keyboard letter but are a different character. Use the IPA character.
| symbol | lookalike | note |
| --- | --- | --- |
| ɡ | g | ɡ is U+0261, not keyboard g U+0067 |
| ˈ | ' | ˈ is U+02C8, not the apostrophe U+0027 or the acute accent U+00B4 |
| θ | ϑ | θ is U+03B8, not the theta symbol U+03D1 |
| ʝ | ɟ | ʝ is U+029D; keyboard j U+006A is the glide, a different symbol |
| ɲ | ñ | ɲ is U+0272, not the letter ñ U+00F1 |
| tʃ | t͡ʃ | write tʃ with no tie bar |
`ɾ` and `r` are two symbols with two meanings: *pero* `<<p|ˈ|e|ɾ|o>>`, *perro* `<<p|ˈ|e|r|o>>`.
`ʝ` (the consonant of *yo*) and `j` (the glide of *bien*) are also two symbols.
## Stress and syllables
A **syllable** has one vowel, its **nucleus**. A glide is not a nucleus. So the number of vowel
symbols in a form is its number of syllables: *bien* `<<b|j|e|n>>` has one, *bueno* `<<b|w|ˈ|e|n|o>>` has two.
Put one `ˈ` directly before the stressed vowel of every word of two or more syllables. Put it
after the consonants that start the syllable: *casa* `<<k|ˈ|a|s|a>>`, not `<<ˈ|k|a|s|a>>`.
For a Spanish word, the spelling gives the stressed syllable. Apply the steps in order:
1. An adverb in *-mente* gets one mark, on *men*: *rápidamente*. Do not mark the stem as well.
2. Else, a vowel with a written accent is stressed: *término*, *jamás*.
3. Else, a word that ends in a vowel, or in *n* or *s* after a vowel, is stressed on the next-to-last
syllable: *casa*, *examen*.
4. Else, the word is stressed on the last syllable: *comer*, *usted*. A final *y* counts as a
consonant: *estoy*.
Count the syllables after you write the glides: *nadie* `<<n|ˈ|a|d|j|e>>` has two.
| word | ipa | syllables | rule |
| --- | --- | --- | --- |
| que | `<<k\|e>>` | 1 | one syllable: no mark |
| sí | `<<s\|i>>` | 1 | one syllable: no mark, also with a written accent |
| término | `<<t\|ˈ\|e\|ɾ\|m\|i\|n\|o>>` | 3 | 2: the vowel with the written accent |
| jamás | `<<x\|a\|m\|ˈ\|a\|s>>` | 2 | 2: the vowel with the written accent |
| casa | `<<k\|ˈ\|a\|s\|a>>` | 2 | 3: ends in a vowel, so the next-to-last syllable |
| examen | `<<e\|k\|s\|ˈ\|a\|m\|e\|n>>` | 3 | 3: ends in n, so the next-to-last syllable |
| comer | `<<k\|o\|m\|ˈ\|e\|ɾ>>` | 2 | 4: ends in another consonant, so the last syllable |
| usted | `<<u\|s\|t\|ˈ\|e\|d>>` | 2 | 4: ends in another consonant, so the last syllable |
| estoy | `<<e\|s\|t\|ˈ\|o\|j>>` | 2 | 4: a final y counts as a consonant |
| nadie | `<<n\|ˈ\|a\|d\|j\|e>>` | 2 | 3: count syllables after the diphthong rule |
| rápidamente | `<<r\|a\|p\|i\|d\|a\|m\|ˈ\|e\|n\|t\|e>>` | 5 | 1: an adverb in -mente has one mark, on men |
| termino | `<<t\|e\|ɾ\|m\|ˈ\|i\|n\|o>>` | 3 | 3: minimal set with término and terminó |
| terminó | `<<t\|e\|ɾ\|m\|i\|n\|ˈ\|o>>` | 3 | 2: minimal set with término and termino |
| hacia | `<<ˈ\|a\|θ\|j\|a>>` | 2 | 3: i is a glide, so two syllables |
| hacía | `<<a\|θ\|ˈ\|i\|a>>` | 3 | 2: the accent on í makes a hiatus |
- A written accent on a one-syllable word gives no mark: *tú*, *él*, *sí*, *té*.
- A function word of two or more syllables gets a mark like any other word: *para* `<<p|ˈ|a|ɾ|a>>`.
- A compound written as one word has one mark: *veintidós* `<<b|e|j|n|t|i|d|ˈ|o|s>>`.
- A transcript that drops a written accent (*esta* for *está*) gives the wrong stress. Use the
correctly accented spelling.
## Diphthongs and hiatus
*a e o* are **strong** vowels. *i u* are **weak** vowels. An unstressed weak vowel next to
another vowel is a glide: `j` for *i* and *y*, `w` for *u*. The glide and the vowel make one
syllable, a **diphthong**.
| word | ipa | note |
| --- | --- | --- |
| bien | `<<b\|j\|e\|n>>` | unstressed i before a vowel is the glide j; one syllable |
| bueno | `<<b\|w\|ˈ\|e\|n\|o>>` | unstressed u before a vowel is the glide w |
| hay | `<<a\|j>>` | y after a vowel at a syllable end is j |
| aun | `<<a\|w\|n>>` | unstressed u after a vowel is w (aun, "even") |
| ruido | `<<r\|w\|ˈ\|i\|d\|o>>` | in ui, the first vowel is the glide |
| ciudad | `<<θ\|j\|u\|d\|ˈ\|a\|d>>` | in iu, the first vowel is the glide |
| después | `<<d\|e\|s\|p\|w\|ˈ\|e\|s>>` | an accent on the strong vowel keeps the diphthong |
| reyes | `<<r\|ˈ\|e\|ʝ\|e\|s>>` | y between vowels is the consonant ʝ |
| hielo | `<<j\|ˈ\|e\|l\|o>>` | word-initial hi before a vowel is j |
| muy | `<<m\|u\|j>>` | y after a vowel at the word end is j; one syllable |
A **hiatus** is two vowels in two syllables. Write both as vowels. This happens with two strong
vowels, with a weak vowel that has a written accent next to a strong vowel, and with two identical vowels.
| word | ipa | note |
| --- | --- | --- |
| país | `<<p\|a\|ˈ\|i\|s>>` | an accent on the weak vowel makes a hiatus |
| día | `<<d\|ˈ\|i\|a>>` | an accent on the weak vowel makes a hiatus |
| aún | `<<a\|ˈ\|u\|n>>` | aún ("still") is two syllables; aun ("even") is one |
| poeta | `<<p\|o\|ˈ\|e\|t\|a>>` | two strong vowels are two syllables |
| leer | `<<l\|e\|ˈ\|e\|ɾ>>` | two identical vowels stay two symbols |
| búho | `<<b\|ˈ\|u\|o>>` | a silent h between the vowels changes nothing |
- A glide is never a vowel, and the word *y* ("and") is the vowel: `<<i>>`.
- The 2010 spelling counts *guion*, *truhan*, *fie*, *lie* and *Sion* as one syllable: no mark.
- An accent on the strong vowel keeps the diphthong (*después*). An accent on the weak vowel
breaks it (*hacía* `<<a|θ|ˈ|i|a>>`).
- In *ui* and *iu*, an accent keeps one syllable: *destruí* `<<d|e|s|t|ɾ|w|ˈ|i>>`.
## Spelling chart
Use this chart to write any word. It is the same for every locale and every speaker.
| spelling | IPA | note |
| --- | --- | --- |
| *a* | `<<a>>` | also with an accent; each symbol is the letter |
| *e* | `<<e>>` | also with an accent; each symbol is the letter |
| *i* | `<<i>>` | also with an accent; each symbol is the letter |
| *o* | `<<o>>` | also with an accent; each symbol is the letter |
| *u* | `<<u>>` | also with an accent; each symbol is the letter |
| *i* unstressed, next to another vowel | `<<j>>` | |
| *u* unstressed, next to another vowel | `<<w>>` | |
| *y* after a vowel at a syllable end | `<<j>>` | |
| *b* | `<<b>>` | both letters give `b` |
| *v* | `<<b>>` | both letters give `b` |
| *d* | `<<d>>` | |
| *t* | `<<t>>` | |
| *p* | `<<p>>` | |
| *f* | `<<f>>` | |
| *l* | `<<l>>` | |
| *w* in loans | `<<w>>` | |
| *c* before *a o u* or a consonant | `<<k>>` | |
| *qu* | `<<k>>` | in *que qui*, the *u* is silent |
| *k* | `<<k>>` | |
| *c* before *e i* | `<<θ>>` | |
| *z* | `<<θ>>` | |
| *cc* before *e i* | `<<k\|θ>>` | |
| *sc* before *e i* | `<<s\|θ>>` | |
| *s* | `<<s>>` | also before a voiced consonant: *mismo* `<<m\|ˈ\|i\|s\|m\|o>>` |
| *ch* | `<<tʃ>>` | |
| *tz* | `<<t\|s>>` | *quetzal* `<<k\|e\|t\|s\|ˈ\|a\|l>>` |
| *tl* | `<<t\|l>>` | |
| *g* before *a o u* or a consonant | `<<ɡ>>` | |
| *gu* before *e i* | `<<ɡ>>` | in *gue gui*, the *u* is silent |
| *gü* before *e i* | `<<ɡ\|w>>` | |
| *g* before *e i* | `<<x>>` | |
| *j* | `<<x>>` | |
| *h* | silent | *hola* `<<ˈ\|o\|l\|a>>` |
| *hu* before a vowel | `<<w>>` | *huevo* `<<w\|ˈ\|e\|b\|o>>` |
| *hi* before a vowel | `<<j>>` | *hielo* `<<j\|ˈ\|e\|l\|o>>` |
| *ll* | `<<ʝ>>` | |
| *y* as a consonant | `<<ʝ>>` | |
| *ñ* | `<<ɲ>>` | |
| *m* | `<<m>>` | |
| *n* | `<<n>>` | see "Nasals" |
| *r* | `<<ɾ>>` or `<<r>>` | see "Tap and trill" |
| *x* | `<<k\|s>>` | see "The letter x" |
## One form for Spain and Latin America
Write `θ` for each spelled *z*, and for each *c* before *e* or *i*. Write spelled *s* as `s`.
The form is the same for every speaker. Most speakers from Spain say `θ`. A Latin American speaker
says the same `θ` as *s* (seseo). The voice gives the accent.
| word | ipa | note |
| --- | --- | --- |
| caza | `<<k\|ˈ\|a\|θ\|a>>` | z is θ |
| casa | `<<k\|ˈ\|a\|s\|a>>` | spelled s is s |
| cielo | `<<θ\|j\|ˈ\|e\|l\|o>>` | c before i is θ |
| hacer | `<<a\|θ\|ˈ\|e\|ɾ>>` | c before e is θ |
| luz | `<<l\|u\|θ>>` | final z is θ |
| escena | `<<e\|s\|θ\|ˈ\|e\|n\|a>>` | sc before e is two symbols, s θ |
| acción | `<<a\|k\|θ\|j\|ˈ\|o\|n>>` | cc before i is two symbols, k θ |
| excelente | `<<e\|k\|s\|θ\|e\|l\|ˈ\|e\|n\|t\|e>>` | xc before e is k s θ |
| quetzal | `<<k\|e\|t\|s\|ˈ\|a\|l>>` | tz is t s, not θ |
**The es-419 input form.** A customer may also write a Latin American form. It has no `θ`:
1. Write `s` for each spelled *z*, and for each *c* before *e* or *i*.
2. Where *s* comes before that *c*, write one `s`: *escena* `<<e|s|ˈ|e|n|a>>`.
3. Where *x* comes before that *c*, write `k` and `s`: *excelente* `<<e|k|s|e|l|ˈ|e|n|t|e>>`.
| word | ipa | note |
| --- | --- | --- |
| caza | `<<k\|ˈ\|a\|s\|a>>` | z is s |
| cielo | `<<s\|j\|ˈ\|e\|l\|o>>` | c before e or i is s |
| escena | `<<e\|s\|ˈ\|e\|n\|a>>` | sc before e or i is one s |
| piscina | `<<p\|i\|s\|ˈ\|i\|n\|a>>` | sc before e or i is one s |
| excelente | `<<e\|k\|s\|e\|l\|ˈ\|e\|n\|t\|e>>` | xc before e or i is k s |
| acción | `<<a\|k\|s\|j\|ˈ\|o\|n>>` | cc before e or i is k s |
Both forms are valid inputs.
## Nasals
Write `n` for the letter *n* and `m` for the letter *m*. Write `ɲ` for *ñ*. Do not write the place of a nasal.
- The letter *n* is `n` before every sound: before `k`, `ɡ` or `x`, before `b` (also spelled *v*),
before `f`, before `ʝ` and before `tʃ`.
- The letter *m* is `m`, also before `b` or `p`.
- The letters *nm* are `n m`: two nasals, one for each letter.
- Only *ñ* is `ɲ`.
| word | ipa | note |
| --- | --- | --- |
| banco | `<<b\|ˈ\|a\|n\|k\|o>>` | n before k is n |
| ambos | `<<ˈ\|a\|m\|b\|o\|s>>` | the letter m is m, also before b |
| año | `<<ˈ\|a\|ɲ\|o>>` | ñ is ɲ |
| conmigo | `<<k\|o\|n\|m\|ˈ\|i\|ɡ\|o>>` | the letters nm are n m, two nasals |
| inmóvil | `<<i\|n\|m\|ˈ\|o\|b\|i\|l>>` | the letters nm are n m, two nasals |
| tengo | `<<t\|ˈ\|e\|n\|ɡ\|o>>` | n before ɡ is n |
| ángel | `<<ˈ\|a\|n\|x\|e\|l>>` | n before x is n |
| enviar | `<<e\|n\|b\|j\|ˈ\|a\|ɾ>>` | the letter n is n, also before b (spelled v) |
| invierno | `<<i\|n\|b\|j\|ˈ\|e\|ɾ\|n\|o>>` | the letter n is n, also before b (spelled v) |
| enfermo | `<<e\|n\|f\|ˈ\|e\|ɾ\|m\|o>>` | n before f is n, as the word is spelled |
| información | `<<i\|n\|f\|o\|ɾ\|m\|a\|θ\|j\|ˈ\|o\|n>>` | n before f is n, as the word is spelled |
| infeliz | `<<i\|n\|f\|e\|l\|ˈ\|i\|θ>>` | a prefix is inside the word, so n before f is still n |
| cónyuge | `<<k\|ˈ\|o\|n\|ʝ\|u\|x\|e>>` | the letter n is n, also before ʝ; only ñ is ɲ |
| honra | `<<ˈ\|o\|n\|r\|a>>` | n before another consonant is n |
| un | `<<u\|n>>` | n at the word end is n, also before a next word that starts with f |
The form does not write `β`, `ð` and `ɣ`, because the position of `b`, `d` and `ɡ` gives them.
The nasal rules stop at the word end. *un beso* is two words, `<<u|n>>` and `<<b|ˈ|e|s|o>>`.
## Tap and trill
- **Tap `ɾ`:** a single *r* between vowels, after a consonant other than *l n s* (except after a
prefix, see the trill), and at a syllable end.
- **Trill `r`:** *rr*, *r* at the word start, *r* after *l n s*, and *r* after a prefix that ends
in a consonant (*sub-*, *ab-*, *ob-*).
| word | ipa | note |
| --- | --- | --- |
| pero | `<<p\|ˈ\|e\|ɾ\|o>>` | single r between vowels is the tap ɾ |
| perro | `<<p\|ˈ\|e\|r\|o>>` | rr is the trill r |
| río | `<<r\|ˈ\|i\|o>>` | r at the word start is the trill |
| alrededor | `<<a\|l\|r\|e\|d\|e\|d\|ˈ\|o\|ɾ>>` | r after l is the trill; r at a syllable end is the tap |
| honra | `<<ˈ\|o\|n\|r\|a>>` | r after n is the trill |
| israelí | `<<i\|s\|r\|a\|e\|l\|ˈ\|i>>` | r after s is the trill |
| teatro | `<<t\|e\|ˈ\|a\|t\|ɾ\|o>>` | r after another consonant in the same syllable is the tap |
| subrayar | `<<s\|u\|b\|r\|a\|ʝ\|ˈ\|a\|ɾ>>` | after the prefix sub-, r is the trill |
## The letter x
- Between vowels and before a consonant, *x* is `<<k|s>>`.
- At the word start, *x* is `s`.
- In *México*, *Oaxaca* and their families, and some other names, *x* is `x`. The word decides.
- In some Mexican names from Nahuatl, *x* is `s`: *Xochimilco* `<<s|o|tʃ|i|m|ˈ|i|l|k|o>>`.
| word | ipa | note |
| --- | --- | --- |
| examen | `<<e\|k\|s\|ˈ\|a\|m\|e\|n>>` | x between vowels is k s |
| texto | `<<t\|ˈ\|e\|k\|s\|t\|o>>` | x before a consonant is k s |
| xenofobia | `<<s\|e\|n\|o\|f\|ˈ\|o\|b\|j\|a>>` | x at the word start is s |
| méxico | `<<m\|ˈ\|e\|x\|i\|k\|o>>` | the word decides; in México, x is x |
| Oaxaca | `<<w\|a\|x\|ˈ\|a\|k\|a>>` | initial Oa in the Oaxaca family is w a |
## English words and brands
Write an English word or a brand with ONE form, the way a monolingual Spanish speaker says it.
This is the English-based adaptation: each English sound becomes the nearest Spanish sound.
Use only the 26 symbols. Never write English sounds (`ɹ`, `æ`, `ə`, `ɪ`, `ʊ`, `ʌ`, English
diphthongs).
Use the spelling chart for a loan that Spanish dictionaries list or that has a Spanish spelling
(*vídeo*, *fútbol*, *chat*, *test*). Use the English adaptation for every other English word or brand.
Put the mark on the vowel that the English word stresses, not where the Spanish spelling rule puts it
(*Netflix* in the next table).
| word | ipa | note |
| --- | --- | --- |
| Microsoft | `<<m\|ˈ\|a\|j\|k\|ɾ\|o\|s\|o\|f>>` | aɪ is a j; the final t is dropped |
| Walmart | `<<w\|ˈ\|a\|l\|m\|a\|ɾ>>` | the final t is dropped |
| okay | `<<o\|k\|ˈ\|e\|j>>` | oʊ is o; eɪ is e j |
| Netflix | `<<n\|ˈ\|e\|t\|f\|l\|i\|k\|s>>` | ɛ is e; ɪ is i |
| Starbucks | `<<e\|s\|t\|ˈ\|a\|ɾ\|b\|a\|k\|s>>` | e before initial s + consonant; ʌ is a; ɹ is ɾ |
| Facebook | `<<f\|ˈ\|e\|j\|s\|b\|u\|k>>` | eɪ is e j; ʊ is u |
| Google | `<<ɡ\|ˈ\|u\|ɡ\|e\|l>>` | spelled oo is u; final ə is e |
### English sounds in Spanish
Each row is one English sound, or one group of English sounds. The note gives the Spanish sound.
| word | ipa | note |
| --- | --- | --- |
| web | `<<w\|e\|b>>` | p b t d k ɡ f s m n l tʃ w stay the same |
| vision | `<<b\|ˈ\|i\|ʝ\|o\|n>>` | v is b; ʒ is ʝ |
| Zoom | `<<s\|u\|m>>` | z is s |
| the | `<<d\|e>>` | ð is d |
| thriller | `<<θ\|ɾ\|ˈ\|i\|l\|e\|ɾ>>` | θ stays θ |
| show | `<<tʃ\|o\|w>>` | ʃ is tʃ; oʊ is o w when stressed and word final |
| jeans | `<<ʝ\|i\|n\|s>>` | dʒ and ʒ are ʝ |
| hashtag | `<<x\|ˈ\|a\|tʃ\|t\|a\|ɡ>>` | h is x |
| shopping | `<<tʃ\|ˈ\|o\|p\|i\|n>>` | ŋ is n |
| burger | `<<b\|ˈ\|u\|ɾ\|ɡ\|e\|ɾ>>` | ɹ is r at the word start and after l n s, else ɾ. ɜ ɹ and ə ɹ are the spelling vowel and ɾ, else e ɾ |
| yes | `<<ʝ\|e\|s>>` | j is ʝ at the word start or between vowels, else j |
| Walmart | `<<w\|ˈ\|a\|l\|m\|a\|ɾ>>` | ɑ ɒ ɔ are the spelling vowel a or o; if neither, a (o for ɔ) |
| Starbucks | `<<e\|s\|t\|ˈ\|a\|ɾ\|b\|a\|k\|s>>` | ʌ is a |
| Amazon | `<<ˈ\|a\|m\|a\|s\|o\|n>>` | æ is a |
| ketchup | `<<k\|ˈ\|e\|tʃ\|a\|p>>` | ə is the last letter of the spelling vowel; spelled u is a; else a when stressed, e when unstressed |
| Facebook | `<<f\|ˈ\|e\|j\|s\|b\|u\|k>>` | eɪ is e j; aɪ is a j; ɔɪ is o j; aʊ is a w |
- Add an unstressed `e` before an initial *s* + consonant: *spa* `<<e|s|p|ˈ|a>>`.
- A final consonant cluster of an English word is as short as a Spanish speaker says it:
*Microsoft* and *Walmart* have no final `t`.
## Conventions
- **Citation form.** Write each word as a speaker says it alone, with care. Write *usted*
`<<u|s|t|ˈ|e|d>>` with its final `d`, and *para* `<<p|ˈ|a|ɾ|a>>` in full.
- **One word at a time.** Do not join words. *y* ("and") is `<<i>>`, never `<<ʝ>>`.
- **Doubled letters.** Write a doubled consonant once: *innovar* `<<i|n|o|b|ˈ|a|ɾ>>`. *cc* before
*e i* is two sounds, `<<k|θ>>`. *ll* and *rr* are single sounds, `ʝ` and `r`.
- **Acronyms.** An acronym said as a word is a word. An acronym said letter by letter joins the
Spanish letter names into one form with one mark, on the last letter name.
- **Digits.** Write the number as words first, then give each word its form.
## Do's and don'ts
Other notations write these symbols for sounds of Spanish words. In a Spanish word, write the
symbol of the second column. For an English word, use the table of English sounds.
| never | write | word | ipa |
| --- | --- | --- | --- |
| ʎ | `<<ʝ>>` | calle | `<<k\|ˈ\|a\|ʝ\|e>>` |
| β | `<<b>>` | abogado | `<<a\|b\|o\|ɡ\|ˈ\|a\|d\|o>>` |
| ð | `<<d>>` | todo | `<<t\|ˈ\|o\|d\|o>>` |
| ɣ | `<<ɡ>>` | agua | `<<ˈ\|a\|ɡ\|w\|a>>` |
| z | `<<s>>` | mismo | `<<m\|ˈ\|i\|s\|m\|o>>` |
| ʃ, ʒ | `<<ʝ>>` | yo | `<<ʝ\|o>>` |
| h | | hola | `<<ˈ\|o\|l\|a>>` |
| ɛ | `<<e>>` | mesa | `<<m\|ˈ\|e\|s\|a>>` |
| ɔ | `<<o>>` | sol | `<<s\|o\|l>>` |
| ə | `<<a>>` | casa | `<<k\|ˈ\|a\|s\|a>>` |
| ɪ | `<<i>>` | sí | `<<s\|i>>` |
| ʊ | `<<u>>` | luz | `<<l\|u\|θ>>` |
| ts | `<<t\|s>>` | quetzal | `<<k\|e\|t\|s\|ˈ\|a\|l>>` |
| aj (one symbol) | `<<a\|j>>` | hay | `<<a\|j>>` |
| ŋ | `<<n>>` | tengo | `<<t\|ˈ\|e\|n\|ɡ\|o>>` |
| ɱ | `<<n>>` | enfermo | `<<e\|n\|f\|ˈ\|e\|ɾ\|m\|o>>` |
| ˌ | | rápidamente | `<<r\|a\|p\|i\|d\|a\|m\|ˈ\|e\|n\|t\|e>>` |
| ː | | leer | `<<l\|e\|ˈ\|e\|ɾ>>` |
## Definitions
- **broad**: a transcription that writes phonemes, not the exact sounds. This page is broad.
- **phoneme**: a sound that tells two words apart, like `p` and `b` in *pala* and *bala*.
- **seseo**: saying *z*, soft *c* and *s* all as *s*. Latin America, the Canary Islands, parts of
Andalusia.
- **es**: the one form for every Spanish locale. **es-ES**: Spain. **es-419**: Latin America.
Note for LLM: as G2P, write the output only as `<<k|ˈ|a|s|a>>`, a word between `<<` and `>>` with `|` between the symbols. In the tables you see a backslash before each `|`. Never write that backslash in your output. It is only for the table display.
Output format:
- Give each word its own <<>>. Use the Spanish inventory above.
- Reply with exactly one <<>> line per word, in the order given, and nothing else — no notes or corrections.
Words to transcribe: