Guide to Pronunciation

The English Language in the Dictionary Index   Spelling

The English language has, of course, both a written form and a spoken form.  Each of these is in a state of continual and inexorable change.  The written language, however, is more stable than the spoken.  With the rapid spread of printing, particularly from the 16th to the 18th centuries, printers and scholars gradually adopted an increasingly fixed set of spelling conventions.  Most of these are adhered to today, though there is still some orthographic variation.  Because the spoken language is more susceptible to change, we find ourselves now with spellings that often reveal more about the history of English than about current pronunciation.  It has become necessary, therefore, for English dictionaries to indicate pronunciation in order to provide an adequate picture of the English vocabulary.

Just as present-day English is not static, it is also not completely uniform.  The pronunciation, vocabulary, and grammar of people living in different areas differ in varying degrees.  Similarly, people who have different levels of education, who hold different sorts of jobs, or who, for one reason or another, move only among certain segments of society, may have distinct forms of speech.  In fact, each person's speech is distinguishable in some ways from that of everyone else; thus, we are able to identify people by the sound of their voices.  However, largely because speech is primarily a form of communication, certain patterns in pronunciation, vocabulary, and usage can be discerned among the members of regional or social groups that regularly communicate with one another.  The more isolated a particular group is from other speakers of the same language, the greater the differences between the speech of that group and the speech of others will tend to become.   As such isolation continues, these differences increase, eventually resulting in distinct dialects.  Given a long enough time and sufficiently limited intercommunication, the speech of a particular group may become so incomprehensible to others that it can be classified as a separate language.  The evolution of Latin into French, Italian, Portuguese, Rumanian, and Spanish is an example of this process.

As geographical and social barriers are overcome, a more widespread mutual comprehensibility becomes possible.  Especially as a result of the technological developments in transportation and communication during the past century, many millions of people are crossing or communicating across dialect boundaries.  We can therefore be assured that slowly and, for the most part, imperceptibly these boundaries will continue to shift; some may disappear and new ones may develop.

During this century researchers into differences in pronunciation, vocabulary, and grammar have been able to discern patterns that define three major geographical dialect areas in the U.S. -- Northern, Southern, and Midland.  Each of these areas also has recognizable subdivisions.  These dialectal differences began with early settlers from Britain who brought their various dialects to different areas along the eastern coast.  With the subsequent isolation from Britain the speech of the settlers in North America gradually diverged from that of the parts of Britain from which they came.  When the population expanded and moved westward, the various dialects were carried westward in an increasingly complex pattern as migration paths crossed and as new lines of communication opened up.  However, the pronunciation of English throughout the U.S. is still based on the three major dialects, which are most clearly defined in the eastern states.  In Canada, which has had closer contact with and more recent settlement from Britain, greater similarity to British speech can be heard with a mixture of some Northern U.S. features.  There are, of course, other aspects of pronunciation, grammar, and vocabulary that arose in or are confined to Canada itself.

There are several broad types or classes of pronunciation variation and these are covered in various ways in a general dictionary.  In each dialect area the significant individual sounds (or phonemes) of the language may be articulated differently from those in other areas.  Many of the features that we perceive as differences in accent fall into this class.  In Southern speech, for example, the vowel of tip or bit is pronounced differently from the same vowel in Northern speech. A Southern speaker seeing the symbol  \ i \  in the pronunciation respelling of a word in this dictionary can turn to the chart of pronunciation symbols and find there the common words tip, active, and banish illustrating the pronunciation of that vowel.  This speaker knows the sound native to his or her own dialect and can use it in the pronunciation of the word in question.  A Northern or Midland speaker will do the same, reproducing the natural and appropriate variety of the sound.  It is not necessary, therefore, that a general dictionary indicate these variations for every word.  In effect, they are covered implicitly by the set of pronunciation symbols used in the book.

Some dialectal differences are the result not simply of variation in the sounds themselves but in the choice of sounds used.  This type of variation is regularly shown in this dictionary.  Research has revealed, for example, that south of a line running irregularly westward through Maryland, northern Virginia, and southern West Virginia, the word creek is usually pronounced \'kre¯k\.  North of this line the more frequent pronunciation is \'krik\, though the spelling pronunciation \'kre¯k\ is often heard as well.  Another such line could be drawn just north of New York City, south along the westward boundary of New Jersey, and northwestward through Pennsylvania.  To the north of this line the great majority of people pronounce greasy as \'gre¯-se¯\, while to the south most people say \'gre¯-ze¯\.  Comparison of a large number of such items makes the boundaries of dialect areas apparent.  For those cases in which the distribution of variants is restricted to a fairly simple pattern within one or two dialect areas, it is possible to label them appropriately, as the entries for great and help where a \Southern\ label appears and for figure where a \Brit\ label indicates a variant heard most frequently in British speech.

A third type of variation that figures widely in this dictionary may be called unpredictable variation.  For many words the distribution of variants is either so widespread (if not random) or so complex in relation to the defined dialect areas that it is impossible to predict accurately which variant or variants a speaker from a particular area might use.  This very common variation is represented in the pronunciation of such words as economic, ration, envelope, and temperature, to name only four out of thousands.  Often when a foreign word is borrowed into English, a number of variants will coexist as people attempt either to reproduce its foreign pronunciation or perhaps to pronounce it according to the spelling, as if it were English in origin.  Thus we hear a number of pronunciation variants for such well-established words as junta and lingerie.

A major problem that arises in editing a dictionary is how to determine the incidence and extent of pronunciation variants.  Since the middle of the 1930's the Merriam-Webster pronunciation editors have been doing their best to solve this problem by carrying out an ongoing program of listening to, recording, and transcribing the pronunciation of educated speakers of English, especially in the United States, Britain, and Canada.  The result of that program is a unique and extensive file of transcriptions that provides the data on which decisions regarding pronunciation are based.

The Merriam-Webster pronunciation file consists primarily of a collection of 3´5 slips of paper (citations), each of which contains a transcription of the pronunciation of a word actually used by someone.  Along with the transcription are included the name of the speaker, additional identifying information, and the date.  These citations are collected by listening to radio, television, and live speech, and in these days of network and satellite broadcasting it is possible to hear a wide range of speakers from all over the English-speaking world.  It is generally inadvisable, however, to transcribe the pronunciation of actors in performance, since they may not be using their natural speech.  The best source of pronunciation is a native speaker of English who can be identified by name and whose geographical and educational background is known.  When most people speak, whether privately or publicly, they concentrate more on the content of what they are saying than on the pronunciation of each word.  This is ideal for the linguist (or pronunciation editor) who is interested in learning how people speak when communicating with others.

When an entry in a Merriam-Webster dictionary is written or revised, the pronunciation citations for that word are reviewed to determine whether it has any pronunciation variants that are sufficiently widespread to warrant inclusion.  One fact that the evidence in our pronunciation file makes apparent, and that is reflected in our dictionaries, is that there is a considerable amount of perfectly acceptable pronunciation variation in the language.  Unless restricted by a regional or other usage label, all of the variants shown in this book fall within the range of acceptable variation.

No system of indicating pronunciation is self-explanatory.  The following discussion sets out the signification and use of the pronunciation symbols and devices in this book, with special attention to those areas where experience has shown that dictionary users may have questions.  More detailed information can be found in the Guide to Pronunciation in Webster's Third New International Dictionary.  The order of symbols discussed below is the same as the order on the page of Pronunciation Symbols, with the exception that the symbols which are not letter characters are listed here first.

\  \ All pronunciation information is printed between reversed virgules.  Pronunciation symbols are printed in roman type and all other information, such as labels and notes, is printed in italics.

\ ' , \ A high-set stress mark precedes a syllable with primary (strongest) stress;
a low-set mark precedes a syllable with secondary (medium) stress; a third level of weak stress requires no mark at all: \'pen-men-,ship\.
Since the nineteenth century the International Phonetics Association has recommended that stress marks precede the stressed syllable, and linguists worldwide have adopted this practice on the basic principle that before a syllable can be uttered the speaker must know what degree of stress to give it.  In accordance with the practice of French phoneticians, no stress marks are shown in the transcription of words borrowed from French whose pronunciations have not been anglicized, as at ancien re´gime and e´meute.

\ - \ Hyphens are used to separate syllables in pronunciation transcriptions.  In actual speech, of course, there is no pause between the syllables of a word.  The placement of these hyphens is based on phonetic principles and may not match the end-of-line divisions indicated by centered dots in boldface entry words.

\ ( ) \ Parentheses are used in pronunciations to indicate that whatever is symbolized between them is present in some utterances but not in others; thus factory \'fak-t(e-)re¯\ is pronounced both \'fak-te-re¯\ and \'fak-tre¯\, industry \'in-(,)des-tre¯\ is pronounced both \'in-des-tre¯\ and \'in-,des-tre¯\.  In some phonetic environments, as in fence \'fen(t)s\ and more \'mo¯(e)r, 'mo¯(e)r\, it may be difficult to determine whether the sound shown in parentheses is or is not present in a given utterance; even the usage of a single speaker may vary considerably.

\ ,; \ Variant pronunciations are separated by commas; groups of variants are separated by semicolons.  The order of variants does not mean that the first is in any way preferable to or more acceptable than the others.  All of the variants in this book, except those restricted by a regional or usage label, are widely used in acceptable educated speech.  If evidence reveals that a particular variant is used more frequently than another, the former will be given first.  This should not, however, prejudice anyone against the second or subsequent variants.  In many cases the numerical distribution of variants is equal but one of them, of course, must be printed first.

\ ¸ \  The obelus, or division sign, is placed before a pronunciation variant that occurs in educated speech but that is considered by some to be questionable or unacceptable.  This symbol is used sparingly and primarily for variants that have been objected to over a period of time in print by commentators on usage, in schools by teachers, or in correspondence that has come to the Merriam-Webster editorial department.  In most cases the objection is based on orthographic or etymological arguments.  For instance, the second variant of cupola \'kyu¯-pe-le, ¸-,lo¯\, though used frequently in speech, is objected to because a is very rarely pronounced \ o¯ \ in English.  The pronunciations
\'feb-ye-,wer-e¯\ and \'feb-e-,wer-e¯\ (indicated simultaneously by the use of parentheses) are similarly marked at the entry for February \¸'feb-(y)e-,wer-e¯, 'feb-re-\, even though they are the most frequently heard pronunciations, because some people insist that both r's should be pronounced.  The obelus applies only to that portion of the transcription which it immediately precedes and not to any other variants following.

\ e \  in unstressed syllables as in banana, collide, abut.  This neutral vowel may be represented orthographically by any of the letters a, e, i, o, u, y, and by many combinations of letters. Unstressed \ e \ often intrudes between a stressed vowel and a following \ l \ or \ r \ though it is not represented in the spelling, as in eel \'e¯(e)l\, wire \'wi¯(e)r\, corn \'ko¯(e)rn\, sour \'sau¯(e)r\.

\'e, ,e\  in stressed syllables as in humdrum, abut.

\ ^e \  immediately preceding \ l \, \ n \, \ m \, \ n \, as in battle, cotton, and one pronunciation of open \'o¯p-^em\ and of and \ ^en \ as in one pronunciation of the phrase lock and key \,la¯k-^en-'ke¯\. The symbol \ ^e \ preceding these consonants does not itself represent a sound.  It signifies instead that the following consonant is syllabic; that is, the consonant itself forms the nucleus of a syllable that does not contain a vowel.
In the pronunciation of some French or French-derived words \ ^e \ is placed immediately after \ l \, \ m \, \ r \ to indicate one nonsyllabic pronunciation of these consonants, as in the French words table ``table,'' prisme ``prism,'' and titre ``title,'' each of which in isolation and in some contexts is a one-syllable word.

\ er \  as in further, merger, bird.  (See the section on \ r \.)  The anglicized pronunciation of the vowel \ œ \ is represented in this book as \e(r)\.  (See the section on \ œ \.)

\'er-, 'e-r\  as in two different pronunciations of hurry.  Most U.S. speakers pronounce \'her-e¯\ with the \ er \ representing the same sounds as in bird \'berd\.  Usually in metropolitan New York and southern England and frequently in New England and the southeastern U.S. the vowel is much the same as the vowel of hum followed by a syllable-initial variety of \ r \.  This pronunciation of hurry is represented as \'he-re¯\ in this book.  Both types of pronunciation are shown for words composed of a single meaningful unit (or morpheme) as in current, hurry, and worry.  In words such as furry, stirring, and purring in which a vowel or vowel-initial suffix is added to a word ending in r or rr (as fur, stir, and purr), the second type of pronunciation outlined above is heard only occasionally and is not shown in this dictionary.

\ a \  as in mat, map, mad, gag, snap, patch.  Some variation in this vowel is occasioned by the consonant that follows it; thus, for some speakers map, mad, and gag have noticeably different vowel sounds.  There is a very small number of words otherwise identical in pronunciation that these speakers may distinguish solely by variation of this vowel, as in the two words can (put into cans; be able) in the sentence ``Let's can what we can.''  However, this distinction is sufficiently infrequent that the traditional practice of using a single symbol is followed in the book.

\ a¯ \  as in day, fade, date, aorta, drape, cape.  In most English speech this is actually a diphthong.  In lowland South Carolina, in coastal Georgia and Florida, and occasionally elsewhere \ a¯ \ is pronounced as a monophthong.  As a diphthong \ a¯ \ has a first element \ e \ or monophtongal \ a¯ \ and a second element \ i \.

\ a¯ \  as in bother, cot, and, with most American speakers, father, cart.  The symbol \ a¯ \ represents the vowel of cot, cod, and the stressed vowel of collar in the speech of those who pronounce this vowel differently from the vowel in caught, cawed, and caller, represented by \ o¯ \. In U.S. speech \ a¯ \ is pronounced with little or no rounding of the lips, and it is fairly long in duration, especially before voiced consonants.  In southern England \ a¯ \ is usually accompanied by some lip rounding and is relatively short in duration.  The vowel \ o¯ \ generally has appreciable lip rounding.  Some U.S. speakers (a perhaps growing minority) do not distinguish between cot--caught, cod--cawed, and collar--caller, usually because they lack or have less lip rounding in the words transcribed with \ o¯ \.  Though the symbols \ a¯ \ and \ o¯ \ are used throughout this book to distinguish the members of the above pairs and similar words, the speakers who rhyme these pairs will automatically reproduce a sound that is consistent with their own speech.
In words such as card and cart most U.S. speakers have a sequence of sounds that we transcribe as \ a¯r \.  Most speakers who do not pronounce \ r \ before another consonant or a pause, however, do not rhyme card with either cod or cawed and do not rhyme cart with either cot or caught.  The pronunciation of card and cart by such speakers, although not shown in this dictionary, would be transcribed as \'ka¯d\ and \'ka¯t\.  Speakers of r-dropping dialects will automatically substitute \ a¯ \ for the transcribed \ a¯r \.  (See the sections on \ a¯ \ and \ r \.)

\ a¯ \  as in father as pronounced by those who do not rhyme it with bother.  The pronunciation of this vowel varies regionally.  In eastern New England and southern England it is generally pronounced farther forward in the mouth than
\ a¯ \ but not as far forward as \ a \.  In New York City and the southeastern U.S. it may have much the same quality as \ a¯ \ but somewhat greater duration.
In areas in which \ r \ is not pronounced before another consonant or a pause, \ a¯ \ occurs for the sequence transcribed in this book as \ a¯r \.  (See the sections on \ a¯ \ and \ r \.)  In these areas \ a¯ \ also occurs with varying frequency in a small group of words in which a in the spelling is followed by a consonant letter other than r and is not preceded by w or wh, as in father, calm, palm, and tomato but not in watch, what, or swap (though \ a¯ \ does sometimes occur in waft). Especially in southern England and, less consistently, in eastern New England
\ a¯ \ occurs in certain words in which \ a \ is the usual American vowel and in most of which the vowel is followed by \ f  \ , \ th \, \ s \, or by \ n \ and another consonant.  The following words and word elements are among the most susceptible to the \ a¯ \ pronunciation.  Where a appears in the spelling more than once, the vowel that may be pronounced \ a¯ \ is marked with a dot.

adva¯nce, adva¯ntage, aft, after, agha¯st, answer, ask, aunt, avala¯nche, bask, basket, bath, behalf, blanch, blast, branch, brass, calf, calve, can't, cask, casket, cast, caste, caster, castle, castor, chaff, chance, chancel, chancellor, chancery, chandler, chant, clasp, class, command, dance, demand, fancy, fast, fasten, flabberga¯st, flask, gasp, ghastly, giraffe, glance, glass, graft, graph,
-graph, grass, grasp, half, halve, lance, last, lath, laugh, mask, mast, master, nasty, pass, past, pastor, path, plant, plaster, prance, raft, rafter, ra¯scal, rasp, raspberry, remand, repast, reprimand, salve (n), sample, sampler, shaft, shan't, slander, slant, staff, stanchion, supplant, task, trance, trans-, va¯ntage, vast

The pronunciation with \ a¯ \ is to be understood as a variant for all of the words discussed in this paragraph, though \ a¯ \ is shown only for those few in which it occurs with especially high frequency.
The symbol \ a¯ \ is also used in the transcription of some foreign-derived words and names. This vowel, as in French patte ``paw'' and chat ``cat,'' is intermediate between \ a \ and \ a¯ \ and is similar in quality to the \ a¯ \ heard in eastern New England.

\ au¯ \  as in now, loud, out.  The initial element of this diphthong may vary from \ a \ to \ a¯ \ or \ a¯ \, the first being more common in Southern and south Midland speech than elsewhere.  In coastal areas of the southern U.S. and in parts of Canada this diphthong is often realized as \ eu¯ \ when immediately preceding a voiceless consonant, as in the noun house and in out.

\ b \  as in baby, rib.

\ ch \  as in chin, nature \'na¯-cher\.  Actually, this sound is  \ t \ + \ sh \.  The distinction between the phrases why choose and white shoes is maintained by a difference in the juncture of the \ t \ and the \ sh \ in each case and the consequent use of different varieties (or allophones) of \ t \.

\ d \  as in did, adder.

\ e \  as in bet, bed, peck.

\'e¯, ,e¯\  in stressed syllables as in beat, nosebleed, evenly, easy.

\ e¯ \  in unstressed syllables, as in easy, mealy.  Though the fact is not shown in this book, some dialects such as southern British and southern U.S. often, if not usually, pronounce \ i \ instead of unstressed \ e¯ \.

\ f \  as in fifty, cuff.

\ g \  as in go, big, gift.

\ h \  as in hat, ahead.

\ hw \  as in whale as pronounced by those who do not have the same pronunciation for both whale and wail.  Most U.S. speakers distinguish these two words as \'hwa¯(e)l\ and \'wa¯(e)l\ respectively, though frequently in the U.S. and usually in southern England \'wa¯(e)l\ is used for both.  Some linguists consider
\ hw \ to be a single sound, a voiceless \ w \.

\ i \  as in tip, banish, active.

\ i¯ \  as in site, side, buy, tripe.  Actually, this sound is a diphthong, usually composed of \ a¯ \ + \ i \ or \ a¯ \ + \ i \.  In Southern speech, especially before a pause or voiced consonant, as in shy and five, the second element \ i \ may not be pronounced.  Chiefly in eastern Virginia, coastal South Carolina, and parts of Canada the diphthong is approximately \'e\ + \ i \ before voiceless consonants, as in nice and write.

\ j \  as in job, gem, edge, join, judge.  Actually, this sound is \ d \ + \ zh \.  Assuming the anglicization of Jeanne d'Arc as \zha¯n-'da¯rk\, the distinction between the sentences They betray John Dark and They betrayed Jeanne d'Arc is maintained by a difference in the juncture of the \ d \ and the \ zh \ in each case and the consequent use of different varieties (or allophones) of \ d \.

\ k \  as in kin, cook, ache.

\ k_ \  as in German ich ``I,'' Buch ``book,'' and one pronunciation of English loch.  Actually, there are two distinct sounds in German; the \ k_ \ in ich is pronounced toward the front of the mouth and the \ k_ \  in Buch is pronounced toward the back.  In English, however, no two words otherwise identical are distinguished by these two varieties of  \ k_ \ , and therefore only a single symbol is necessary.  In English speech the front variety of  \ k_ \  is produced automatically to accompany a front vowel, such as \ e \ or \ i \, and the back variety to accompany a back vowel, such as \ a¯ \ or \ u¯ \.

\ l \  as in lily, pool.  In words such as battle and fiddle the \ l \ is a syllabic consonant.  (See the section on \ ^e \ above.)

\ m \  as in murmur, dim, nymph.  In pronunciation variants of some words, such as open and happen, \ m \ is a syllabic consonant.  (See the section on \ ^e \ above.)

\ n \  as in no, own.  In words such as cotton and sudden, the \ n \ is a syllabic consonant.  (See the section on \ ^e \ above.)

\ ⁿ \  indicates that a preceding vowel or diphthong is pronounced with the nasal passages open, as in French un bon vin blanc \œⁿ-bo¯ⁿ-vaⁿ-bla¯ⁿ\ ``a good white wine.''

\ n \  as in sing \'sin\, singer \'sin-er\, finger \'fin-er\, ink \'ink\.  In some contexts \ n \ may be a syllabic consonant.  (See the section on \ ^e \ above.)

\ o¯ \  as in bone, know, beau.  Especially in positions of emphasis, such as when it is word final or when as primary stress, \ o¯ \ tends to become diphthongal, moving from \ o¯ \ toward a second element \ u¯ \.  In southern England and in some U.S. speech, particularly in the Philadelphia area and in the Pennsylvania-Ohio-West Virginia border area, the first element is often approximately \ e \.  In coastal South Carolina, Georgia, and Florida stressed
\ o¯ \ is often monophthongal when final, but when a consonant follows it is often a diphthong moving from \ o¯ \ to  \ e \.  In this book the symbol \ o¯ \ represents all of the above variants.

\ o¯ \  as in saw, all, gnaw, caught.  (See the section on \ a¯ \.)

\ œ \  as in French boeuf ``beef,'' German Ho¯lle ``hell.''  This vowel, which occurs only in foreign-derived terms and names, can be approximated by attempting to pronounce the vowel \ e \ with the lips moderately rounded as for the vowel \ u¯ \.  This vowel is often anglicized as the \ er \ of bird by those who do not ``drop their r's'' or as the corresponding vowel of bird used by those who do (see the section on \ r \).  Where this anglicization is shown, it is represented as \e(r)\.

\ œ¯¯ \  as in French feu ``fire,'' German Ho¯hle ``hole.''  This vowel, which occurs primarily in foreign-derived terms and names, can be approximated by attempting to pronounce a monophthongal vowel \ a¯ \ with the lips fully rounded as for the vowel \ u¯ \.  This vowel also occurs in Scots and thus is used in the pronunciation of guidwillie, mainly restricted to Scotland.

\ o¯i \  as in coin, destroy.  In some Southern speech, especially before a consonant in the same word, the second element may disappear or be replaced by \ e \.  Some utterances of drawing and sawing have a sequence of vowel sounds identical to that in coin, but because drawing and sawing are analyzed by many as two-syllable words they are transcribed with a parenthesized hyphen: \'dro¯(-)in\, \'so¯(-)in\.

\ p \  as in pepper, lip.

\ r \  as in red, rarity, car, beard.  In some dialects, especially those of the southeastern U.S., eastern New England, New York City, and southern England, \ r \ is not pronounced when another consonant or a pause follows immediately.  This is often, if somewhat misleadingly, referred to as r-dropping.  In these dialects r is pronounced as a nonsyllabic \ e \ when it occurs in these positions or there may be no sound corresponding to the r; thus beard, corn, and assured may be pronounced as \'bied\, \'ko¯en\, and \e-'shu¯ed\ or, usually with some lengthening of the vowel sound, as \'bid\, \'ko¯n\, and \e-'shu¯d\.  In car, card, and cart those who do not pronounce \ r \ generally have a vowel which we would transcribe as \ a¯ \, usually pronounced with some lengthening and without a following \ e \.  (See the sections on \ a¯ \ and \ a¯ \.)  The stressed vowel of bird and hurt in r-dropping speech is similar to the vowel used by r-keepers in the same words but without the simultaneous raising of the center and/or tip of the tongue.  In the U.S. most speakers of r-dropping dialects will pronounce \ r \ before consonants in some words or in some contexts.  Because it is determined by the phonetic context, r-dropping is not explicitly represented in this dictionary; speakers of r-dropping dialects will automatically substitute the sounds appropriate to their own speech.

\ s \  as in source, less.

\ sh \  as in shy, mission, machine, special.  Actually this is a single sound, not two.  When the two sounds \ s \ and \ h \ occur in sequence, they are separated by a hyphen in this book, as in grasshopper \'gras-,ha¯p-er\.

\ t \  as in tie, attack, late, later, latter.  In some contexts, as when a stressed or unstressed vowel precedes and an unstressed vowel or \ ^el \ follows, the sound represented by t or tt is pronounced in much American speech the same as the sound represented by d or dd in similar contexts.  Thus, the pairs ladder and latter, leader and liter, parody and parity are often homophones.  In such instances this dictionary shows \ d \ at the end of a syllable for those words spelled with d or dd (\'lad-er\, \'le¯d-er\, \'par-ed-e¯\) and \ t \ at the end of a syllable for those with t or tt (\'lat-er\, \'le¯t-er\, \'par-et-e¯\).

\ th \  as in thin, ether.  Actually, this is a single sound, not two.  When the two sounds \ t \ and \ h \ occur in sequence they are separated by a hyphen in this book, as in knighthood \'ni¯t-,hu¯d\.

\ _t_h \  as in then, either, this.  Actually, this is a single sound, not two.  The basic difference between \ th \ and \ _t_h \ is that the former is pronounced without and the latter with vibration of the vocal cords.

\ u¯ \  as in rule, youth, union \'yu¯n-yen\, few \'fyu¯\.

\ u¯ \  as in pull, wood, book, curable \'kyu¯r-e-bel\, fury \'fyu¯(e)r-e¯\.

\ u \  as in German fu¯llen ``to fill,'' hu¯bsch ``handsome.'' This vowel, which occurs only in foreign-derived terms and names, can be approximated by attempting to pronounce the vowel \ i \ with the lips moderately rounded as for the vowel \ u¯ \.

\ u¯ \  as in French rue ``street,'' German fu¯hlen ``to feel.''  This vowel, which occurs only in foreign-derived terms and names, can be approximated by attempting to pronounce the vowel \ e¯ \ with the lips fully rounded as for the vowel \ u¯ \.

\ v \  as in vivid, invite.

\ w \  as in we, away.  In some words having final \ (,)o¯ \, as follow, \(,)yu¯\, as value, or \(,)u¯\, as statue, an unstressed variant \ e \ or \ ye \ may occur, especially before a consonant or a pause, as in \ 'fa¯l-ed\ or \'val-yed\, and a variant \e-w\ or \ye-w\ occurs before vowels, as in \'fa¯l-e-win\ or \'val-ye-win\.  These variants are transcribed \e(-w)\ or \ye(-w)\ at the entry word.

\ y \  as in yard, young, cue \'kyu¯\, curable \'kyu¯r-e-bel\, few \'fyu¯\, fury \'fyu¯(e)r-e\, union \'yu¯n-yen\.  The sequences \lyu¯\, \syu¯\, and \zyu¯\ in the same syllable, as in lewd, suit, and presume, are common in southern British speech but are rare in American speech and only \ lu¯ \, \ su¯ \, and \ zu¯ \ are shown in this dictionary.
In English \ y \ does not occur at the end of a syllable after a vowel.  In a few words of French origin whose pronunciation has not been anglicized, a postvocalic \ y \ is transcribed, as in mille-feuille \me¯l-fœy\ and in rouille \'ru¯-e¯, F ru¯y\.  The sound represented is the consonantal \ y \ of yard.

\ ^y \  indicates that during the articulation of the preceding consonant the tongue has substantially the position it has for the articulation of the \ y \ of yard, as in French digne \de¯n^y\ ``worthy.''  Thus \ ^y \ does not itself represent a sound but rather modifies the preceding symbol.

\ z \  as in zone, raise.

\ zh \  as in vision, azure \'azh-er\.  Actually, this is a single sound, not two.  When the two sounds \ z \ and \ h \ occur in sequence, they are separated by a hyphen in this book, as in hogshead \'ho¯gz-,hed, 'ha¯gz-\.