Sandhi

Sandhi is any of a wide variety of sound changes that occur at morpheme or word boundaries. Examples include fusion of sounds across word boundaries and the alteration of one sound depending on nearby sounds or the grammatical function of the adjacent words. Sandhi belongs to morphophonology.
Sandhi occurs in many languages, e.g. in the phonology of Indian languages. Many dialects of British English show linking and intrusive R.
Tone sandhi in particular defines tone changes affecting adjacent words and syllables. This is a common feature of many tonal languages such as Burmese and Chinese.

Types

Internal and external sandhi

Sandhi can be either:

internal, at morpheme boundaries within words, such as syn- + pathy: sympathy, or
external, at word boundaries, such as the pronunciation "tem books" for ten books in some dialects of English. The linking process of some dialects of English is a kind of external sandhi, as are French liaison and Italian raddoppiamento fonosintattico.

It may be extremely common in speech, but sandhi is typically ignored in spelling, as is the case in English. Sandhi is, however, reflected in the orthography of Sanskrit, Sinhala, Telugu, Marathi, Pali and some other Indian languages, as with Italian in the case of compound words with lexicalised syntactic gemination.
External sandhi effects can sometimes become morphologised as in Tamil and, over time, turn into consonant mutations.

Tone sandhi

Most tonal languages have tone sandhi in which the tones of words alter according to certain rules. An example is the behavior of Mandarin Chinese; in isolation, tone 3 is often pronounced as a falling-rising tone. When a tone 3 occurs before another tone 3, however, it changes into tone 2, and when it occurs before any of the other tones, it is pronounced as a low falling tone with no rise at the end.
An example occurs in the common greeting 你好 , which is in practice pronounced. The first word is pronounced with tone 2, but the second is unaffected.

Examples

Celtic languages

In Celtic languages, the consonant mutation sees the initial consonant of a word change according to its morphological or syntactic environment.
Following are some examples from Breton, Irish, Scottish Gaelic, and Welsh:

Breton	Welsh	Irish	Scottish Gaelic	Gloss
gwreg	gwraig	bean	bean*	woman/wife
bras	mawr	mór	mòr	big
ar wreg vras	y wraig fawr	an bhean mhór	a' bhean mhòr	the big woman
kazh	cath	cat	cat	cat
e gazh	ei gath	a chat	a chat	his cat
he c'hazh	ei chath	a cat	a cat	her cat
o c'hazh	eu cath	a gcat	an cat	their cat

Portuguese

When two words belonging to the same phrase are pronounced together, or two morphemes are joined in a word, the last sound in the first may be affected by the first sound of the next, either coalescing with it, or becoming shorter, or being deleted. This affects especially the sibilant consonants, and the unstressed final vowels.

Consonant sandhi

As was mentioned above, the dialects of Portuguese can be divided into two groups, according to whether syllable-final sibilants are pronounced as postalveolar consonants or as alveolar. At the end of words, the default pronunciation for a sibilant is voiceless,, but in connected speech the sibilant is treated as though it were within a word :

If the next word begins with a voiceless consonant, the final sibilant remains voiceless ; bons tempos .
If the next word begins with a voiced consonant, the final sibilant becomes voiced as well /z, ʒ/; bons dias .
If the next word begins with a vowel, the final sibilant is treated as intervocalic, and pronounced ; bons amigos .

When two identical sibilants appear in sequence within a word, they reduce to a single consonant. For example, nascer, desço, excesso, exsudar are pronounced with by speakers who use alveolar sibilants at the end of syllables, and disjuntor is pronounced with by speakers who use postalveolars. But if the two sibilants are different they may be pronounced separately, depending on the dialect. Thus, the former speakers will pronounce the last example with, whereas the latter speakers will pronounce the first examples with if they are from Brazil or if from Portugal. This applies also to words that are pronounced together in connected speech:

sibilant + /s/, e.g., as sopas: either ;
sibilant + /z/, e.g., as 'zonas: either ;
sibilant + /ʃ/, e.g., as chaves: always ;
sibilant + /ʒ/, e.g., os g'enes: always .
Vowel sandhi

Normally, only the three vowels /ɐ/, /i/ or /ɨ/, and /u/ occur in unstressed final position. If the next word begins with a similar vowel, they merge with it in connected speech, producing a single vowel, possibly long. Here, "similar" means that nasalization can be disregarded, and that the two central vowels /a, ɐ/ can be identified with each other. Thus,

/aa, aɐ, ɐa, ɐɐ/ → ; toda a noite or , nessa altura or .
/aɐ̃, ɐɐ̃/ → appears only in this environment.
/ii, iĩ/ → , ĩ ; de idade .
/ɨɨ/ → ; fila de espera .
/uu, uũ/ → , ũ ; todo o dia .

If the next word begins with a dissimilar vowel, then /i/ and /u/ become approximants in Brazilian Portuguese :

/i/ + V → ; durante o curso , mais que um .
/u/ + V → ; todo este tempo do objeto .

In careful speech and in with certain function words, or in some phrase stress conditions, European Portuguese has a similar process:

; se a vires , mais que um .
; todo este tempo , do objeto .

But in other prosodic conditions, and in relaxed pronunciation, EP simply drops final unstressed and /u/, though this is subject to significant dialectal variation:

durante o curso , este inquilino .
todo este tempo , disto há muito .

Aside from historical set contractions formed by prepositions plus determiners or pronouns, like à/dà, ao/do, nesse, dele, etc., on one hand and combined cliticpronouns such as mo/ma/mos/mas, and so on, on the other, Portuguese spelling does not reflect vowel sandhi. In poetry, however, an apostrophe may be used to show elision such as in d'água.

German dialects

In various German dialects or the spoken Standard German one can find phonological processes that can be analysed as Sandhi. For example some varieties of Central Hessian show a vowel length alternation where, if the same long vowel were else to repeat in two consecutive syllables, the vowel is shortened/reduced in the first, but maintained in the second. Examples are for HG hingehen or for HG kein einziger.

English

In English phonology, rhotic sandhi can be seen in non-rhotic dialects, when a word ends in a vowel followed by /r/, and the next word starts with a vowel as well, a sound will be inserted between the word, see for example, in Standard Southern British English "law and order" pronounced as , "America and
China" pronounced as

French

French liaison and enchaînement can be considered forms of external sandhi.
In enchaînement, a word-final consonant, when followed by a word that starts with a vowel, is articulated as though it is part of the following word. For example, sens is pronounced and unique is pronounced ; sens unique is pronounced.
Liaison is a similar phenomenon, applicable to words ending in a consonant that was historically pronounced but that, in Modern French, is normally silent when occurring at the end of a phrase or before another consonant. In some circumstances, when the following word starts with a vowel, the consonant may be pronounced, and in that case is articulated as if part of the next word. For example, deux frères is pronounced with a silent, and quatre hommes is pronounced, but deux hommes is pronounced.

Japanese

In Japanese phonology, sandhi is primarily exhibited in rendaku and conversion of つ or く to a geminate consonant, both of which are reflected in spelling – indeed, the っ symbol for gemination is morphosyntactically derived from つ, and voicing is indicated by adding two dots as in か／が,, making the relation clear. It also occurs much less often in renjō, where, most commonly, a terminal on one morpheme results in an being added to the start of the next morpheme, as in 天皇: てん＋おう → てんのう, meaning "emperor"; that is also shown in the spelling.

Korean

has sandhi which occurs in the final consonant or consonant cluster, such that a morpheme can have two pronunciations depending on whether or not it is followed by a vowel. For example, the root 읽, meaning ‘read’, is pronounced before a consonant, as in 읽다, but is pronounced like before vowels, as in 읽으세요, meaning ‘please read’. Some roots can also aspirate following consonants, denoted by the letter ㅎ in the final consonant. This causes 다 to become in 않다, ‘to not be’.

Tamil

Tamil 'punarchi' or sandhi has been rigorously and exhaustively documented in Tamil grammar texts since the early centuries AD. As modern Tamil is strongly characterised by diglossia: there are two separate registers varying by speech context, a high register and a low one. This in turn presents two corresponding domains for forming Sandhi. Tamil employs Sandhi for certain morphological and syntactic structures.