CS Knowledge Base

The Extended Speech Assessment Methods Phonetic Alphabet (X-SAMPA) is a variant of SAMPA developed in 1995 by John C. Wells, professor of phonetics at University College London.^[1] It is designed to unify the individual-language SAMPA alphabets, and extend SAMPA to cover the entire range of characters in the 1993 version of International Phonetic Alphabet (IPA). The result is a SAMPA-inspired remapping of the IPA into 7-bit ASCII.

SAMPA was devised as a hack to work around the inability of text encodings to represent IPA symbols. Later, as Unicode support for IPA symbols became more widespread, the necessity for a separate, computer-readable system for representing the IPA in ASCII decreased. However, X-SAMPA is still useful as the basis for an input method for true IPA.

SAMPA was developed in the late 1980s in the European Commission-funded ESPRIT project 2589 "Speech Assessment Methods" (SAM)—hence "SAM Phonetic Alphabet"—in order to facilitate email data exchange and computational processing of transcriptions in phonetics and speech technology. As many symbols as possible were taken over from the IPA; where this is not possible, other signs that are available were used, e.g. [@] for schwa (IPA [ə]), [2] for [ø] (the vowel sound found in French deux ('two')), and [9] for [œ] (the vowel sound found in French neuf ('nine')).

SAMPA, as originally standardized, is a partial encoding of the IPA. The first version of SAMPA was the union of the sets of phoneme codes for Danish, Dutch, English, French, German and Italian; later versions extended SAMPA to cover other European languages. Since SAMPA is based on phoneme inventories, each SAMPA table is valid only in the language it was created for. In order to make this IPA encoding technique universally applicable, X-SAMPA was developed, which provides one single table without language-specific differences.

Language-specific SAMPA representations have been developed for the sounds of the following languages:^{[citation needed]}

X-SAMPA uses the ASCII notation below for IPA representation.

• The IPA symbols that are ordinary lower case letters have the same value in X-SAMPA as they do in the IPA.
• X-SAMPA uses backslashes as modifying suffixes to create new symbols. For example, s is a distinct sound from s\, which are phonologically related in many ways. Such use of the backslash character can be a problem, since many programs interpret it as an escape character for the character following it. For example, such X-SAMPA symbols do not work in EMU, so backslashes must be replaced with some other symbol (e.g., an asterisk: '*') when adding phonemic transcription to an EMU speech database. The backslash has no fixed meaning.
• X-SAMPA uses the underscore character to join following diacritics to the symbols they modify, except for ~ (nasalization), = (syllabicity), and ` (retroflexion and rhotacization), for which the underscore is optional.
• The underscore character is also used to encode the IPA tiebar: k_p codes for /k͡p/.
• The numbers _1 to _6 are reserved diacritics as shorthand for language-specific tone numbers.
• The IETF language tags registry has assigned fonxsamp as the subtag for text transcribed in X-SAMPA.^[3]

Lower-case symbols

X-SAMPA	IPA	IPA image	Description	Examples
a	a		open front unrounded vowel	French dame [dam]
b	b		voiced bilabial plosive	English bed [bEd], French bon [bO~]
b_<	ɓ		voiced bilabial implosive	Sindhi ɓarʊ [b_<arU]
c	c		voiceless palatal plosive	Hungarian latyak ["lQcQk]
d	d		voiced alveolar plosive	English dig [dIg], French doigt [dwa]
d`	ɖ		voiced retroflex plosive	Swedish hord [hu:d`]
d_<	ɗ		voiced alveolar implosive	Sindhi ɗarʊ [d_<arU]
e	e		close-mid front unrounded vowel	French blé [ble]
f	f		voiceless labiodental fricative	English five [faIv], French femme [fam]
g	ɡ		voiced velar plosive	English game [geIm], French longue [lO~g]
g_<	ɠ		voiced velar implosive	Sindhi ɠəro [g_<@ro]
h	h		voiceless glottal fricative	English house [haUs]
h\	ɦ		voiced glottal fricative	Czech hrad [h\rat]
i	i		close front unrounded vowel	English be [bi:], French oui [wi], Spanish si [si]
j	j		palatal approximant	English yes [jEs], French yeux [j2]
j\	ʝ		voiced palatal fricative	Greek γειά [j\a]
k	k		voiceless velar plosive	English skip [skIp], Spanish carro ["karo]
l	l		alveolar lateral approximant	English lay [leI], French mal [mal]
l`	ɭ		retroflex lateral approximant	Svealand Swedish sorl [so:l`]
l\	ɺ		alveolar lateral flap	Wayuu püülükü [pM:l\MkM]
m	m		bilabial nasal	English mouse [maUs], French homme [Om]
n	n		alveolar nasal	English nap [n{p], French non [nO~]
n`	ɳ		retroflex nasal	Swedish hörn [h2:n`]
o	o		close-mid back rounded vowel	French veau [vo]
p	p		voiceless bilabial plosive	English speak [spik], French pose [poz], Spanish perro ["pero]
p\	ɸ		voiceless bilabial fricative	Japanese fuku [p\M_0kM]
q	q		voiceless uvular plosive	Arabic qasbah ["qQs_Gba]
r	r		alveolar trill	Spanish perro ["pero]
r`	ɽ		retroflex flap	Bengali gari [gar`i:]
r\	ɹ		alveolar approximant	English red [r\Ed]
r\`	ɻ		retroflex approximant	Malayalam വഴി ["v@r\`i]
s	s		voiceless alveolar fricative	English seem [si:m], French session [sE"sjO~]
s`	ʂ		voiceless retroflex fricative	Swedish mars [mas`]
s\	ɕ		voiceless alveolo-palatal fricative	Polish świerszcz [s\v'ers`ts`]
t	t		voiceless alveolar plosive	English stew [stju:], French raté [Ra"te]
t`	ʈ		voiceless retroflex plosive	Swedish mört [m2t`]
u	u		close back rounded vowel	English boom [bu:m], Spanish su [su]
v	v		voiced labiodental fricative	English vest [vEst], French voix [vwa]
v\ (or P)	ʋ		labiodental approximant	Dutch west [v\Est]/[PEst]
w	w		labial-velar approximant	English west [wEst], French oui [wi]
x	x		voiceless velar fricative	Scots loch [lOx] or [5Ox]; German Buch, Dach; Spanish caja, gestión
x\	ɧ		voiceless palatal-velar fricative	Swedish sjal [x\A:l]
y	y		close front rounded vowel	French tu [ty] German über ["y:b6]
z	z		voiced alveolar fricative	English zoo [zu:], French azote [a"zOt]
z`	ʐ		voiced retroflex fricative	Mandarin Chinese rang [z`aN]
z\	ʑ		voiced alveolo-palatal fricative	Polish źrebak ["z\rEbak]

Upper-case symbols

X-SAMPA	IPA	IPA image	Description	Example
A	ɑ		open back unrounded vowel	English father ["fA:D@(r\)] (RP and Gen.Am.)
B	β		voiced bilabial fricative	Spanish lavar [la"Ba4]
B\	ʙ		bilabial trill	Medumba mʙʉ [mB\}_H]
C	ç		voiceless palatal fricative	German ich [IC], English human ["Cjum@n] (broad transcription uses [hj-])
D	ð		voiced dental fricative	English then [DEn]
E	ɛ		open-mid front unrounded vowel	French même [mE:m], English met [mEt] (RP and Gen.Am.)
F	ɱ		labiodental nasal	English emphasis ["EFf@sIs] (spoken quickly, otherwise uses [Emf-])
G	ɣ		voiced velar fricative	Greek γωνία [Go"nia]
G\	ɢ		voiced uvular plosive	Inuktitut nirivvik [niG\ivvik]
G\_<	ʛ		voiced uvular implosive	Mam ʛa [G\_<a]
H	ɥ		labial-palatal approximant	French huit [Hit]
H\	ʜ		voiceless epiglottal fricative	Agul мехӀ [mEH\]
I	ɪ		near-close front unrounded vowel	English kit [kIt]
I\	ᵻ		near-close central unrounded vowel (non-IPA)	Polish ryba [rI\bA]
J	ɲ		palatal nasal	Spanish año ["aJo], English canyon ["k{J@n] (broad transcription uses [-nj-])
J\	ɟ		voiced palatal plosive	Hungarian egy [EJ\]
J\_<	ʄ		voiced palatal implosive	Sindhi ʄaro [J\_<aro]
K	ɬ		voiceless alveolar lateral fricative	Welsh llaw [KaU]
K\	ɮ		voiced alveolar lateral fricative	Mongolian долоо [tOK\O:]
L	ʎ		palatal lateral approximant	Italian famiglia [fa"miLLa], Castilian: llamar [La"mar]
L\	ʟ		velar lateral approximant	Korean 달구지 [t6L\gudz\i]
M	ɯ		close back unrounded vowel	Korean 음식 [M:ms\_hik_}]
M\	ɰ		velar approximant	Spanish fuego ["fweM\o]
N	ŋ		velar nasal	English thing [TIN]
N\	ɴ		uvular nasal	Japanese さん san [saN\]
O	ɔ		open-mid back rounded vowel	American English off [O:f]
O\	ʘ		bilabial click
P (or v\)	ʋ		labiodental approximant	Dutch west [PEst]/[v\Est], allophone of English phoneme /r\/
Q	ɒ		open back rounded vowel	RP lot [lQt]
R	ʁ		voiced uvular fricative	German rein [RaIn]
R\	ʀ		uvular trill	French roi [R\wa]
S	ʃ		voiceless postalveolar fricative	English ship [SIp]
T	θ		voiceless dental fricative	English thin [TIn]
U	ʊ		near-close back rounded vowel	English foot [fUt]
U\	ᵿ		near-close central rounded vowel (non-IPA)	English euphoria [jU\"fO@r\i@]
V	ʌ		open-mid back unrounded vowel	Scottish English strut [str\Vt]
W	ʍ		voiceless labial-velar fricative	Scots when [WEn]
X	χ		voiceless uvular fricative	Klallam sχaʔqʷaʔ [sXa?q_wa?]
X\	ħ		voiceless pharyngeal fricative	Arabic ح ḥāʾ [X\A:]
Y	ʏ		near-close front rounded vowel	German hübsch [hYpS]
Z	ʒ		voiced postalveolar fricative	English vision ["vIZ@n]

Other symbols

X-SAMPA	IPA	IPA image	Description	Example
.	.		syllable break
"	ˈ		primary stress
%	ˌ		secondary stress	American English pronunciation [pr\@%nVn.si"eI.S@n]
' (or _j)	ʲ		palatalized	Russian Земля (Earth) [z'I"ml'a] or [z_jI"ml_ja]
:	ː		long
:\	ˑ		half long	Estonian differentiates three vowel lengths
-			separator	Polish trzy [t-S1] vs. czy [tS1] (affricate)
@	ə		schwa	English arena [@"r\i:n@]
@\	ɘ		close-mid central unrounded vowel	Chuvash пӗррехинче [p@\rrEXints\E]
@`	ɚ		r-coloured schwa	American English color ["kVl@`]
{	æ		near-open front unrounded vowel	English trap [tr\{p]
}	ʉ		close central rounded vowel	Swedish sju [x\}:]; AuE/NZE boot [b}:t]
1	ɨ		close central unrounded vowel	Welsh tu [t1], American English rose's ["r\oUz1z]
2	ø		close-mid front rounded vowel	Danish købe ["k2:b@], French deux [d2]
3	ɜ		open-mid central unrounded vowel	English nurse [n3:s] (RP) or [n3`s] (Gen.Am.)
3\	ɞ		open-mid central rounded vowel	Irish tomhail [t3\:l']
4	ɾ		alveolar flap	Spanish pero ["pe4o], American English better ["bE4@`]
5	ɫ		velarized alveolar lateral approximant; also see _e	English milk [mI5k], Portuguese livro ["5iv4u]
6	ɐ		near-open central vowel	German besser ["bEs6], Australian English mud [m6d]
7	ɤ		close-mid back unrounded vowel	Estonian kõik [k7ik], Vietnamese mơ [m7_M]
8	ɵ		close-mid central rounded vowel	Swedish buss [b8s]
9	œ		open-mid front rounded vowel	French neuf [n9f], Danish drømme [dR9m@]
&	ɶ		open front rounded vowel	Swedish skörd [x\&d`]
?	ʔ		glottal stop	Cockney English bottle ["bQ?o]
?\	ʕ		voiced pharyngeal fricative	Arabic ع ʿayn [?\Ajn]
*			undefined escape character, SAMPA's "conjunctor"
/	/		(a) French vowel archiphonemes or indeterminacies (b) delimiter of phonemic transcriptions	maison /mE/zO~/
<	⟨		begin nonsegmental notation, e.g., SAMPROSA[4]
<\	ʢ		voiced epiglottal fricative	Siwi arˤbˤəʢa (four) [ar_?\b_?\@<\a]
>	⟩		end nonsegmental notation
>\	ʡ		epiglottal plosive	Archi гӀарз (complaint) [>\arz]
^	ꜛ		upstep
!	ꜜ		downstep
!\	ǃ		postalveolar click	Zulu iqaqa (polecat) [i:!\a:!\a]
|	|		minor (foot) group
|\	ǀ		dental click	Zulu icici (earring) [i:|\i:|\i]
||	‖		major (intonation) group
|\|\	ǁ		alveolar lateral click	Zulu xoxa (to converse) [|\|\O:|\|\a]
=\	ǂ		palatal click	Taa ǂnûm [=\~u_Fm]
-\	‿		linking mark

Diacritics

X-SAMPA	IPA	IPA image	Description
_"	̈		centralized
_+	̟		advanced
_-	̠		retracted
_/	̌		rising tone
_R
_0	̥		voiceless
_<			implosive (IPA uses separate symbols for implosives)
= (or _=)	̩		syllabic
_>	ʼ		ejective
_?\	ˤ		pharyngealized
_\	̂		falling tone
_F
_^	̯		non-syllabic
_}	̚		no audible release
`	˞		rhotacization in vowels, retroflexion in consonants (IPA uses separate symbols for consonants, see t` for an example)
~ (or _~)	̃		nasalization
_A	̘		advanced tongue root
_a	̺		apical
_B	̏		extra low tone
_B_L	᷅		low rising tone
_c	̜		less rounded
_d	̪		dental
_e	̴		velarized or pharyngealized; also see 5
_f	↘		global fall[citation needed]
_G	ˠ		velarized
_H	́		high tone
_H_T	᷄		high rising tone
_h	ʰ		aspirated
_j (or ')	ʲ		palatalized
_k	̰		creaky voice
_L	̀		low tone
_l	ˡ		lateral release
_M	̄		mid tone
_m	̻		laminal
_N	̼		linguolabial
_n	ⁿ		nasal release
_O	̹		more rounded
_o	̞		lowered
_q	̙		retracted tongue root
_r	↗		global rise [citation needed]
_R_F	᷈		rising falling tone
_r	̝		raised
_T	̋		extra high tone
_t	̤		breathy voice
_v	̬		voiced
_w	ʷ		labialized
_X	̆		extra-short
_x	̽		mid-centralized

The SAM PROSodic Alphabet (SAMPROSA) is a system for prosodic transcription,^[4] developed by Dafydd Gibbon, Daniel J. Hirst, and William J. Barry as part of the SAM Prosody Working Group.^[5][1] SAMPROSA is composed of the prosodic symbols of SAMPA along with a SAM System for Intonation Transcription (SAMSINT).^[5] There is some overlap between the segmental symbols of SAMPA and the prosodic symbols of SAMPROSA, however, and so X-SAMPA separates out SAMPROSA notation from SAMPA by the use of the < and > tier escape symbols. The IPA's alternative way of encoding tones (ȅ, e̋, etc) also derives from the same SAMPROSA notational system: using B,L,M,H,T; R,F following a _.

Prosodic symbols and diacritics

X-SAMPA	IPA	IPA image	Description
<^> (or ^)	ꜛ		upstep
<!> (or !)	ꜜ		downstep
</> (or <R>)	↗		global rise
<\> (or <F>)	↘		global fall
<T>	˥		extra high tone
<H>	˦		high tone
<M>	˧		mid tone
<L>	˨		low tone
<B>	˩		extra low tone
<BT> (or <LH>, etc)	˩˥		rising tone
<TB> (or <HL>, etc)	˥˩		falling tone
<HT>	˧˥		high rising tone
<BL>	˩˧		low rising tone
<HTH> (or <MHM>, etc)	˧˥˧		rising falling tone

Such tone letters as above are represented in Unicode through combining modifier characters rendered as ligatures. For example, ˧˥ (<HT>) is represented by ˧ (<H>) and ˥ (<T>) directly adjacent, with no separating space. Occasionally, language-specific tone numbers (<1>, <2>, etc) are used instead of their explicit counterparts (eg. <TT>, <M>), but such usage across languages is not standardized as a part of X-SAMPA.

• Comparison of ASCII encodings of the International Phonetic Alphabet
• SAMPA chart for English
• List of phonetics topics
• BABEL Speech Corpus

• Computer-coding the IPA: A proposed extension of SAMPA Archived 2011-07-09 at the Wayback Machine
• X-SAMPA to IPA to CXS converter
• Web-based translator for X-SAMPA documents. Produces Unicode text, XML text, PostScript, PDF, or LaTeX TIPA.
• Z-SAMPA, a backward-compatible extension of X-SAMPA sometimes used for conlangs
• SAMPA computer readable phonetic alphabet
• Phonemic notation of English in SAMPA
• Converter from (German) written text to SAMPA and IPA (Ajax-application)
• IPA-SAMPA Converter and IPA-SAMPA chart