A Monthly Column for Word Lovers
Add (or Subtract!) Your Voice
English has eight pairs of consonant sounds that are naturally aligned on the basis of voice—which is to say, for each pair, your mouth parts are in the same configuration for pronunciation and the main thing that distinguishes one consonant from its pair mate is whether your vocal cords are vibrating at the moment of articulation. To feel this difference, place your fingers on your throat and make a /sssssss/ sound like a snake. Now, produce a steady /zzzzzz/, like a bee buzzing. The hum, and the vibration you feel in your throat with /zzzzzz/, is voicing. These eight pairs of voiced and voiceless sounds are scattered across the lexicon of English in some ways that give strong support to Henry Fowler's astute observation that relations among words in English
come to us from our forefathers as an odd jumble and plainly show that the language has not been neatly constructed by a master builder who could create each part to do the exact work required of it, neither overlapped or overlapping; far from that, its parts have had to grow as they could.
On the other hand, some English sound pairs behave in quite regular ways and offer up clues about the relationships among words. Far better, then, to focus on the interesting patterns of these consonant pairs than to lament the randomness of their legacy. Let's begin with the sounds typically represented by the pair S and Z.
Of these two letters (and always remember: letters do not consistently represent the same sound), Z is by far the less complicated because its presence in a word is nearly always a clue that you should gear up for a voiced alveolar fricative (in other words; hissing while your vocal cords are vibrating). The presence of S in a word, however, might be a placeholder for a voiceless alveolar fricative (hissing, no vibration), or it may in fact be pronounced as /z/. How do we know which it is? For native speakers it's pretty much second nature. If you look at the esses in the first sentence of this paragraph, you see the following patterns and the (simplified) rules that govern them:
S in letters is pronounced as /z/ because it follows a voiced consonant (r).
S in less is pronounced as /s/ because there are two of them.
S in because, presence and always is pronounced as /z/ because it follows a voiced vowel.
The general takehome: if your vocal cords are already vibrating and an S comes along in a word, don't stop the vibe; pronounce the S as /z/. If consecutive esses come along, the first one should stop your vocal cords from vibrating and the second one is pronounced unvoiced. Are these rules universal? Of course not, since we're talking about English. There's vase (pronounced with either /s/ or /z/) and there's buses, in which the first S is pronounced as /s/, even though it would make more sense to do this if you use the alternate spelling busses. Learners of English may want to go further down this rabbit hole to explore all the intricacies of the S and Z pair here.
A similar pattern emerges with D (voiced) and T (unvoiced), with a few twists. The easier pattern is the pronunciation of D as /t/ in verb inflections when the terminal sound of the verb root is not voiced—as in voiced! Speakers follow the same rule here: if your vocal cords are vibrating, keep them vibrating for a terminal D (braided, framed, sawed), but if they're not vibrating, don't start them just for the inflectional ending: change your voiced velar plosive /d/ to an unvoiced one, /t/ (flipped, sucked, passed). But of course if the E of the inflection is voiced, resulting in another syllable, then you have already revved up your vocal cords and you might as well continue: stated, muted. To add further confusion but also to confirm the rule, consider the pronunciation of the past tense blessed, /blest/, and the adjective blessed, /blesəd/. The same pattern applies to marked (past tense of mark, /markt/) and marked (two-syllable adjective, /markəd).
A less confusing and thus more reassuring pattern emerges between the pair of labiodental fricatives, voiced V and voiceless F. Their great virtue is that their presence in a word is usually an indication of their expected pronunciation. Their peculiar twist is that they show up in word pairs that are cognate but that typically represent different parts of speech. Examples are legion: give/gift, halve/half, heave/heft, shrive/shrift, cleave/cleft, prove/proof, grieve/grief, believe/belief, bereave/bereft, save/safe, shelve/shelf, weave/weft. For some combinations, the semantic connection is less obvious because of the way words have changed in meaning, but the relationship is the same: for example, thrive and thrift (thrift originally meant "vigorous growth"). In other cases, one word of the pair has essentially become archaic, leaving its partner to fend for itself in English, such as rive, the more or less obsolete partner of rift). The pleasing pattern you may have noticed is that a verb with V very often has a cognate noun or adjective with F; but then you've got sieve (noun) and sift (verb).
The V/F pair of sounds is completely dependable in irregular plurals: leaf/leaves, roof/rooves, calf/calves, half/halves, thief/thieves, loaf/loaves, knife/knives. These words are all English and Germanic natives. The noun beef is not an English native, but a loaner from French, on which the pressure of conformity has resulted in the unusual plural beeves.
Though it's always a good idea to speak well of the dead, we may often be tempted to berate our ancestors for the ambiguities they bequeathed us with one other pair of consonant sounds, the voiced and unvoiced interdental fricatives, for which we have not two separate letters, but instead a single digraph, TH. Speakers must apply rules to determine in a given word whether TH is voiced (as in this) or unvoiced (as in thick).
Our ancestors up through Middle English had a slightly easier go with these two sounds because they spelled them with separate letters: the th in this was spelled with the letter eth, ð, and the th in thick was spelled with thorn, þ. Today, happily, this pair of consonant sounds also behaves regularly and parts of speech, along with conventions of spelling, clue us up as to which sound to use at the end of a word most of the time. In all of the following pairs, the terminal e (in verbs) is a clue to voiced TH, while its absence (in nouns or adjectives) means the sound is voiceless: bathe/bath, breathe/breath, clothe/cloth, loathe/loath, sheathe/sheath, soothe/sooth, teethe/teeth, wreathe/wreath.
The force of the pattern is such that it applies to some pairs with no spelling difference, such as mouth (verb, voiced) and mouth (noun, unvoiced), but it doesn't exert its force on smooth (voiced for all parts of speech), or birth (unvoiced for all parts of speech, though oddly voiced, sometimes, in the inflection birthing). If this maddening inconsistency is troubling to you, there may be some comfort in knowing that voiced TH (the sound in this), is no longer productive in English. If you hand somebody a new object with a label saying "Zeth", they will never say /zɛð/. We do not make new words with that sound.
Is there any more method to this madness? Unfortunately not. It may be some comfort for learners of English that variations in pronunciation of the sounds noted above are common and expected, and indeed, part of the charm of a "foreign accent" is hearing one of these consonant sounds when you expect another.