Voiced consonants are produced with the vocal cords vibrating.
Voiceless consonants: the vocal cords are not vibrating.)
In English, you see an effect of this difference in the pronunciation of the regular noun plural ending -s. English speakers assume that because the plural -s is spelled "s", that it is always pronounced [ s ] (as in mats, cuffs, cups, ducks, etc.).
However, this is not always the case. In the following examples, the the plural -s is pronounced [ z ]: fogs, clubs, rods, gloves, etc.
In English the "voicing" quality of the plural ending -s depends on the voicing of the preceding consonant sound. In other words, when the last consonant sound in the noun root is a voiceless consonant, the plural ending is also voiceless: [ s ]. Otherwise, the plural ending is voiced: [ z ], as the following chart indicates:
|Voicing quality:||Example consonant sounds:||Plural ending is prounounced:||Example words:|
|voiceless||p t k f||[ s ]||ducks|
|voiced||b d g v||[ z ]||logs|
(Note: this process is called "assimilation" because the consonant -s (because it is voiceless) is more similar to another voiceless consonant sound that precedes it. When the plural -s is pronounced [ z ] (which is a voiced sound), it is more similar to another voiced sound.)
Voicing assimilation in Latvian
We also have voicing assimilation in Latvian, but it works in the opposite direction. Take a look at the following nouns, which all have a nominative singular masculine ending -s:
koks 'tree', prāts 'mind, sense', acs 'eye', auss 'ear'
In all of the above examples, the ending -s is pronounced as [ s ], which is what one might expect, since it immediately follows a voiceless consonant sound such as [ k ], [ t ], [ ʦ ] or [ s ].
However, the ending -s is still pronounced as [ s ] in the following example words:
sirds 'heart', zirgs 'horse', labs 'good'
Instead, it is the final consonant of the root word which is devoiced. Thus, the final sounds in these words are pronounced as follows: [ ts ] sirds 'heart', [ ks ] zirgs 'horse', [ ps ] labs 'good'.
So, you see that, in Latvian, it is the following sound which influences the voicing of the preceding sound.
This also occurs when a prefix is added to the front of a root word - the final consonant sound of the prefix takes on the voicing quality of the following consonant sound. Look at the following example words, which illustrate all of these points:
|Spelling of Affected Sound||Pronunciation of Affected Sound||Example word||Translation||2nd Example word||Translation|
|In the following examples, a voiceless sound (like p t k c s š ) is followed by a voiced sound (like: b, d, g, v, z, ž) and, thus, is also pronounced as a voiced sound.
||[ b ]
||to bury, inter
||[ d ]
||[ g ]
||[ ʣ ]
||five years in length/age
||[ z ]
||[ ʒ ]
||In the following examples, a voiced sound (like: b, d, g, v, z, ž) is followed by a voiceless sound (like p t k c s š ) and, thus, is also pronounced as a voiceless sound.
||[ p ]
||[ t ]
||[ k ]
||[ s ]
||to blow away
(Note: although this voicing assimilation is found in numerical compounds, such as trešdiena 'Wednesday, lit. third day', pusdeviņi 'half-past eight', it normally does not occur in regular compound words. Thus, there is no voicing assimilation in compounds such as grīdsega 'carpet, floor covering', pusdzīvs 'half-dead', etc.)
Note: English s and z are slightly different: they are produced by bringing the front of the tongue close to the bony ridge just behind the teeth. They are said to have an alveolar place of articulation.
Compare this to the sounds represented by š, ž, č and dž (phonetically: [ ʃ ], [ ʒ ] [ ʧ ] and [ ʤ ] respectively). These four sounds are produced by bringing the front of the tongue close to hard palate, just behind the bony ridge which is just behind the teeth. They are said to have an alveo-palatal place of articulation.
When these two groups of sounds are put next to each other (by adding a prefix or suffix to a root word), the s or z adjust by taking on the alveo-palatal place of articulation of the š, ž, č or dž. This is illustrated by the examples which follow:
|s||[ ʃ ]||pusčetri||half-past three||visšaurākais||narrowest|
|z||[ ʒ ]||izžūt||to dry out||aizdžinkstēt||to whiz away|
|z||[ ʃ ]||uzšūt||to sew onto||[ z ] takes on the voicelessness of the š which follows it; it also takes on the alveopalatal place of articulation of š and is, thus, pronounced identically to š (i.e. as [ ʃ ])|
|s||[ ʃ ]||mežs||forest||[ s ] takes on the alveo-palatal place of articulation of the preceding ž, and, thus, is pronounced like a š (i.e. as [ ʃ ])|
|ž||[ ʃ ]||mežs||forest||the final consonant of the root noun (i.e. ž - pronounced [ ʒ ]) takes on the voicelessness of the following [ s ] suffix, and, thus, is pronounced pronounced like a š (i.e. as [ ʃ ])|