"Frequency Effects in the Lexical Diffusion of Phonological Change". Betty S. Phillips, Indiana State U. [email protected] LSA Summer Institute Workshop on Variation, Gradience, and Frequency in Phonology, 6-8 July 2007, Stanford U. Lexical Diffusion. - PowerPoint PPT Presentation
"Frequency Effects in "Frequency Effects in the Lexical Diffusion of the Lexical Diffusion of Phonological Change" Phonological Change" Betty S. Phillips, Indiana Betty S. Phillips, Indiana State U. State U. [email protected][email protected]LSA Summer Institute Workshop LSA Summer Institute Workshop on Variation, Gradience, and on Variation, Gradience, and Frequency in Phonology, 6-8 Frequency in Phonology, 6-8 July 2007, Stanford U. July 2007, Stanford U.
Transcript
"Frequency Effects in the Lexical "Frequency Effects in the Lexical Diffusion of Phonological Diffusion of Phonological
Change"Change" Betty S. Phillips, Indiana State U. Betty S. Phillips, Indiana State U.
[email protected]@indstate.eduLSA Summer Institute Workshop on LSA Summer Institute Workshop on
Variation, Gradience, and Frequency in Variation, Gradience, and Frequency in Phonology, 6-8 July 2007, Stanford U.Phonology, 6-8 July 2007, Stanford U.
Lexical DiffusionLexical Diffusion
Definition:Definition:
Chen and Wang (1975): “a Chen and Wang (1975): “a phonological rule gradually extends phonological rule gradually extends its scope of operation to a larger and its scope of operation to a larger and larger portion of the lexicon, until all larger portion of the lexicon, until all relevant items have been transformed relevant items have been transformed by the process.”by the process.”
That lexical diffusion is sporadicThat lexical diffusion is sporadic
That it always affects the most frequent That it always affects the most frequent words first.words first.
That lexically diffused changes show no That lexically diffused changes show no clear phonetic conditioningclear phonetic conditioning
That phonetically regular changes are That phonetically regular changes are productive, whereas lexically diffused productive, whereas lexically diffused changes are not.changes are not.
That lexical diffusion affects only That lexical diffusion affects only phonetically abrupt changes associated phonetically abrupt changes associated with lexical rather than postlexical ruleswith lexical rather than postlexical rules
That word frequency is independent of That word frequency is independent of word class.word class.
That analogy and borrowing suffice to That analogy and borrowing suffice to account for lexical diffusionaccount for lexical diffusion
That age of acquisition or discourse factors That age of acquisition or discourse factors account for patterns of lexical diffusion. account for patterns of lexical diffusion.
That frequency effects are found over a That frequency effects are found over a population of speakers but not within population of speakers but not within individual speakers.individual speakers.
That either lexical diffusionThat either lexical diffusion is the diffusion is the diffusion of a completed sound changeof a completed sound change OR OR "all phonological change starts with lexical diffusion and most ends up Neogrammarian, given enough time"
MisunderstandingMisunderstanding #1 #1
That lexical diffusion is "sporadic" That lexical diffusion is "sporadic"
E.g., Hinskens (1998: 169): "Although variable by E.g., Hinskens (1998: 169): "Although variable by nature, Neogrammarian sound change is pre-nature, Neogrammarian sound change is pre-eminently 'systematic and recurrent' i.e. regular, eminently 'systematic and recurrent' i.e. regular, as there are no lexical exceptions. Lexically as there are no lexical exceptions. Lexically diffuse and lexicalized sound change bring about diffuse and lexicalized sound change bring about 'non-systematic but recurrent' facts (Lloret 1997), 'non-systematic but recurrent' facts (Lloret 1997), which are only partially regular, hence which are only partially regular, hence 'sporadic'."'sporadic'."
One clear way in which lexical diffusion One clear way in which lexical diffusion is not at all random: is not at all random:
Word Frequency EffectsWord Frequency Effects
"Rarely-used words drag behind; very "Rarely-used words drag behind; very frequently used ones hurry ahead. frequently used ones hurry ahead. Exceptions to the sound laws are formed Exceptions to the sound laws are formed in both groups." -- Schuchardt (1885: 58)in both groups." -- Schuchardt (1885: 58)
““A A cursorycursory glance at the newspapers glance at the newspapers suggests that suggests that adulteryadultery is on the increase in is on the increase in this century. If you think this century. If you think slavery slavery has been has been abolished, go and look at the abolished, go and look at the factoryfactory at the at the end of the road. Every mother will tell you end of the road. Every mother will tell you that that nursery nursery schools are a mixed schools are a mixed blessing.” (Aitchison 2003)blessing.” (Aitchison 2003)
Which retain the medial schwa, and which Which retain the medial schwa, and which lose it?lose it?
My ISU class project results:My ISU class project results:
WordWord Celex - raw Celex - raw frequencyfrequency
Schwa-lessSchwa-less Schwa-fulSchwa-ful
everyevery 97889788 2222 44
factoryfactory 10641064 1818 88
nurserynursery 580580 1919 77
slaveryslavery 134134 1010 1616
adulteryadultery 107107 1515 1111
cursorycursory 3232 99 1717
Modern English final /t,d/ deletion Modern English final /t,d/ deletion (Bybee 2002)(Bybee 2002)
Env.Env. More likely to deleteMore likely to delete (Celex frequency) (Celex frequency)
Less likely to deleteLess likely to delete (Celex frequency) (Celex frequency)
-ld-ld told (1763)told (1763) held (765)held (765)
-lt-lt felt (1449)felt (1449) built (456)built (456)
-ft/-st-ft/-st left (1503)left (1503) lost (759)lost (759)
Stress shift in -ate verbsStress shift in -ate verbsStress patterns C$C(C)-Stress patterns C$C(C)-ateate disyllables in Gimson's disyllables in Gimson's 1988 [1991] 1988 [1991] English Pronouncing DictionaryEnglish Pronouncing Dictionary
FrequencyFrequency(Celex)(Celex)
ConservativeConservative⁄ —⁄ —
VariableVariable InnovativeInnovative — ⁄— ⁄
0-70-7
filtrate...filtrate...
XX
9-36 9-36 mandate...mandate...
XX
39-66639-666dictate...dictate...
XX
Misunderstanding #2Misunderstanding #2
That lexically diffused changes always That lexically diffused changes always affect the most frequent words first. affect the most frequent words first.
E.g., E.g., Krug (2003): “It has become a Krug (2003): “It has become a linguistic commonplace that high-linguistic commonplace that high-frequency words and constructions tend to frequency words and constructions tend to lead phonological change. . . .”lead phonological change. . . .”
Middle English Middle English Ormulum Ormulum of /ø:/ > /e:/ of /ø:/ > /e:/ (Phillips 1984: 328)--Verbs(Phillips 1984: 328)--Verbs
1st example: Stress Shift x 1st example: Stress Shift x // > > / / xx
UNCHANGED NOUNSUNCHANGED NOUNS
exchangeexchange
excuseexcuse
exemptexempt
exhaustexhaust
expressexpress
Ave. Freq. = Ave. Freq. = 22.022.0
CHANGED NOUNSCHANGED NOUNS
exciseexcise
exploitexploit
exportexport
extractextract
Ave. Freq. = 2.5Ave. Freq. = 2.5
Summary: Development of Noun-Verb Summary: Development of Noun-Verb Homographs (Phillips 2006)Homographs (Phillips 2006)
05
10152025303540
dis-
ex-
es-
pre- de
- a- re-
con-
/com
-
Unshifted
Diatonic
Misunderstanding #3Misunderstanding #3
That lexically diffused changes show no That lexically diffused changes show no clear phonetic conditioning clear phonetic conditioning (Kiparsky (Kiparsky 1995).1995).
Blevins (2004: 260): "lexical diffusion Blevins (2004: 260): "lexical diffusion (without clear phonetic conditioning) is the (without clear phonetic conditioning) is the diffusion of a completed sound change.”diffusion of a completed sound change.”
knew, duringknew, during, new, new 11.311.3 53.053.0
Misunderstanding #4Misunderstanding #4
That phonetically regular changes are That phonetically regular changes are productive, whereas lexically diffused productive, whereas lexically diffused changes are notchanges are not
Bakken 2001: " Bakken 2001: " A phonetically regular A phonetically regular process is characterized by general process is characterized by general productivity, and such productivity seems to be productivity, and such productivity seems to be incompatible with lexical exceptions, i.e. lexical incompatible with lexical exceptions, i.e. lexical diffusion. Lexical diffusion and phonetic diffusion. Lexical diffusion and phonetic regularity are therefore truly mutually regularity are therefore truly mutually exclusive.” exclusive.”
Labov, Yaeger, and Steiner (1972)Labov, Yaeger, and Steiner (1972)
“ “We find no strong evidence for lexical diffusion in We find no strong evidence for lexical diffusion in the (æh) patterns of Detroit and Buffalo and the (æh) patterns of Detroit and Buffalo and Chicago.”. . . .Chicago.”. . . .
Despite some initial oscillations the (æh) word Despite some initial oscillations the (æh) word class seems to move upward as a whole, with class seems to move upward as a whole, with fine phonetic conditioning in the process. There fine phonetic conditioning in the process. There is some indication that is some indication that the word the word madmad is lower is lower than its phonetic class would justifythan its phonetic class would justify for for several speakers. . . .several speakers. . . . the the low position of low position of madmad as compared to as compared to bad, adsbad, ads, etc. , seems to be , etc. , seems to be lexically determined.lexically determined. " (93 – my bold)." (93 – my bold).
MisunderstandingMisunderstanding #5 #5
That lexical diffusion affects only That lexical diffusion affects only phonetically abrupt changes phonetically abrupt changes associated with lexical rather than associated with lexical rather than postlexical rules postlexical rules (Labov 1994; (Labov 1994; Kiparsky 1995).Kiparsky 1995).
Reduction across word boundariesReduction across word boundaries(By definition "post-lexical")(By definition "post-lexical")
Alba (2003): Alba (2003): la la 'the' or 'the' or una una 'a' + vowel-initial nouns 'a' + vowel-initial nouns (e.g., (e.g., iglesia iglesia 'church', 'church', hija hija 'daughter') in 20 hours 'daughter') in 20 hours of tape-recorded interviews with 20 native of tape-recorded interviews with 20 native speakers of Spanish. speakers of Spanish.
Strings with high ratio frequency underwent hiatus Strings with high ratio frequency underwent hiatus resolution 87 per cent of the time, compared to resolution 87 per cent of the time, compared to 48 per cent for strings with low ratio frequency. 48 per cent for strings with low ratio frequency. That is, That is, high string frequency positively high string frequency positively impacted vowel reductionimpacted vowel reduction..
"Don't""Don't"
Phonetic reduction ranges from initial /do/ Phonetic reduction ranges from initial /do/ to /to /ɾoɾo/ to // to /ɾəɾə/ to // to /əə/ depending on the / depending on the frequency of the phrases in which it frequency of the phrases in which it appears. --Bybee and Scheibman (1999) appears. --Bybee and Scheibman (1999) and Scheibman (2000)and Scheibman (2000)
cf. "I don't know/think/have (to)/want/like" cf. "I don't know/think/have (to)/want/like" vs. "I don't need/follow/make contact"vs. "I don't need/follow/make contact"
Misunderstanding #6Misunderstanding #6
That Word Frequency is independent of That Word Frequency is independent of Word Class in the lexical diffusion of a Word Class in the lexical diffusion of a sound change. sound change.
Modern Dutch diphthongization Modern Dutch diphthongization of WGmc /of WGmc /î/ î/
"We do not find a relationship between token "We do not find a relationship between token frequency and diphthongization....[H]ighly frequency and diphthongization....[H]ighly frequent items such as frequent items such as krijgen, tijd krijgen, tijd (>200) (>200) do not behave differently. . . from items do not behave differently. . . from items such as such as vijg vijg (0) and (0) and vijl vijl (2). And among (2). And among the items showing diphthongs rather often, the items showing diphthongs rather often, such as such as vrijdag, vrijen, vrij, fijn vrijdag, vrijen, vrij, fijn and and vijfvijf, , both low and high frequencies occur."both low and high frequencies occur."
Function words as a special caseFunction words as a special case
Certain sound changes tend to affect word categories with low sentence stress, such as function words, but this does not correlate necessarily with word frequency.
/an, on/ > /on, om/ in /an, on/ > /on, om/ in Pastoral CarePastoral Care (OE, late 9th c.) (OE, late 9th c.)(Phillips 2006)(Phillips 2006)
/ø:/ > /e:/ in the /ø:/ > /e:/ in the OrmulumOrmulum ( ca. 1180 a.d.) ( ca. 1180 a.d.)(Phillips 1984; 2006)(Phillips 1984; 2006)
BY WORD CLASSBY WORD CLASS BY WORD FREQUENCYBY WORD FREQUENCY
% <e>% <e> % <e>% <e>
NumeralsNumerals 00 1-101-10 6262
NounsNouns 2828 11-2011-20 6666
VerbsVerbs 6767 21-3021-30 4545
AdjectivesAdjectives 7070 31-6031-60 4343
Adverbs & Adverbs & Function Function words words
100100 61-90 61-90 2828
91-40091-400 2121
Other Word Classes
Word frequency effects have been shown to be clearest inside of word classes –since in production word class is accessed before the word form.
Specificially, the retrieval of a word's Specificially, the retrieval of a word's syntactic informationsyntactic information precedesprecedes that of that of phonological informationphonological information by approx. by approx. 40 msec40 msec. . -- Turennout et al. (1998)-- Turennout et al. (1998)
Misunderstanding #7Misunderstanding #7
That analogy and borrowing suffice to That analogy and borrowing suffice to account for lexical diffusion account for lexical diffusion (Kiparsky (Kiparsky 1995; Janda and Joseph 2003).1995; Janda and Joseph 2003).
Attribution of Lexical Diffusion effects to Attribution of Lexical Diffusion effects to Analogy & BorrowingAnalogy & Borrowing
““[D]iffusionary effects in the spread of [D]iffusionary effects in the spread of phonological change through the lexicons phonological change through the lexicons of speakers... are actually of speakers... are actually epiphenomenalepiphenomenal, being the result of , being the result of already-needed mechanisms of already-needed mechanisms of analogical changeanalogical change and and dialect dialect borrowingborrowing." ."
----Janda and Joseph (2003)Janda and Joseph (2003)
The Problem with Lexical Diffusion as The Problem with Lexical Diffusion as Analogical ChangeAnalogical Change
Analogical changesAnalogical changes affect the affect the LEAST LEAST FREQUENTFREQUENT words first--"where memory words first--"where memory fails"-- e.g., fails"-- e.g., cows, dragons, aliens BUT cows, dragons, aliens BUT mice, geese, teeth; mice, geese, teeth; RRegular verbs egular verbs (correspond, irrigate, (correspond, irrigate, elucidateelucidate) vs. irregular verbs () vs. irregular verbs (be, go, seebe, go, see))
whereas, whereas, Lexical DiffusionLexical Diffusion oftenoften affects affects the the MOST FREQUENTMOST FREQUENT words first. words first.
Lexical StorageLexical Storage
Bybee, J. 2001. Bybee, J. 2001. Usage-Based PhonologyUsage-Based Phonology. p. 22:. p. 22:
BorrowingBorrowing
Lexical diffusionists "see sound change . . Lexical diffusionists "see sound change . . . as change affecting the sound in certain . as change affecting the sound in certain words and then diffusing gradually to other words and then diffusing gradually to other words in the lexicon. . . . This is like words in the lexicon. . . . This is like 'dialect borrowing', but with some words 'dialect borrowing', but with some words borrowing from others in the same dialect. borrowing from others in the same dialect. It constitutes a different outlook on the It constitutes a different outlook on the transition problem."transition problem." (Campbell 1998: 199) (Campbell 1998: 199)
But surely,But surely,
all change involves borrowingall change involves borrowing: :
“[W]e have no criteria for determining “[W]e have no criteria for determining absolutely that there is an axiomatic absolutely that there is an axiomatic distinction between sound change and distinction between sound change and borrowing (or contact change) because.. borrowing (or contact change) because.. all changes must arise from contact all changes must arise from contact between speakersbetween speakers.” -- Milroy (1992: 88) .” -- Milroy (1992: 88)
Wolfram and Schilling-Estes (2003):Wolfram and Schilling-Estes (2003):
a change becomes propagated a change becomes propagated throughthrough social groups social groups, ,
through through phonologicalphonological and and morphological environmentsmorphological environments, and, and
through the through the lexiconlexicon. .
All of these happen All of these happen simultaneouslysimultaneously, , and even within individuals variation and even within individuals variation is to be expected. is to be expected.
"[T]he notion of "[T]he notion of variability variability . . . applies to . . . applies to both both intra-speakerintra-speaker and and inter-speaker inter-speaker variation. In other words, an individual variation. In other words, an individual speaker will go through a speaker will go through a period of period of fluctuationfluctuation between the old and new between the old and new variant, and speakers within a given variant, and speakers within a given speech community will show speech community will show variation variation from speaker to speakerfrom speaker to speaker with respect to with respect to the use of the new and old variant" --the use of the new and old variant" --Wolfram and Schilling-Estes 2003: 717. Wolfram and Schilling-Estes 2003: 717.
As for changes that clearly imitate an As for changes that clearly imitate an external dialect:external dialect:
With regard to word frequency, there is With regard to word frequency, there is no differenceno difference between "borrowed" between "borrowed" changes and internal changes. changes and internal changes.
In fact, the spread of a sound change from In fact, the spread of a sound change from one dialect to another apparently follows one dialect to another apparently follows the the same direction of diffusionsame direction of diffusion as the as the imitated dialect. E.g.,imitated dialect. E.g.,
Spread of broad /a:/ to the U.S.Spread of broad /a:/ to the U.S.
fashionable import from Britain in the fashionable import from Britain in the eighteenth or nineteenth century.eighteenth or nineteenth century.
A few wordsA few words whose innovative whose innovative pronunciations were pronunciations were not borrowednot borrowed ( (haft; haft; aster; rascal; vantage; ranch, aster; rascal; vantage; ranch, stanchion; Alexander, Flanders, stanchion; Alexander, Flanders, slander)slander),, but but only oneonly one ((alasalas)) that that underwent the underwent the shift in the Eastern U.S. but not in shift in the Eastern U.S. but not in EnglandEngland.-- Phillips 1989.-- Phillips 1989
Supports Trudgill’s (1986: 58) observation Supports Trudgill’s (1986: 58) observation that during accommodation people “modify that during accommodation people “modify their pronunciation of their pronunciation of particular wordsparticular words,... ,... with some words being affected before with some words being affected before others” , which implies that the spread of a others” , which implies that the spread of a sound change from one group to another sound change from one group to another must of necessity proceed via lexical must of necessity proceed via lexical diffusiondiffusion, affecting some words before , affecting some words before others. others.
Second point: Second point:
Of the subset of (already frequent) Of the subset of (already frequent) words that had the innovative vowel words that had the innovative vowel in British English, in British English, Eastern U.S. Eastern U.S. English borrowed the most English borrowed the most frequent.frequent.
Eastern U.S. Words in Eastern U.S. Words in [æ]/[a:] for RP [a:] [æ]/[a:] for RP [a:]
Unchanged Unchanged
Words with [Words with [æ]æ]
Freq. of Freq. of Unchanged Unchanged Words with [Words with [æ]æ]
Ave. Freq. of Words Ave. Freq. of Words with Borrowed [a:]with Borrowed [a:]
Since Since borrowingborrowing (i.e., imitation of other (i.e., imitation of other speakers) is speakers) is inherent to the spreadinherent to the spread of an of an innovation from innovation from speaker to speakerspeaker to speaker
and since and since lexical diffusionlexical diffusion is is inherent in inherent in the spreadthe spread of an innovation from of an innovation from one one word to anotherword to another
BOTH BOTH ARE INHERENT TO SOUND ARE INHERENT TO SOUND CHANGE--not at odds with it.CHANGE--not at odds with it.
AND:AND:
Since the Since the spreadspread of all changes of all changes depends on the similaritiesdepends on the similarities shared shared by lexical entries,by lexical entries,
AnalogyAnalogy, too, is , too, is endemic to sound endemic to sound change.change.
Misunderstanding #8:Misunderstanding #8:
That age of acquisition or discourse That age of acquisition or discourse factors account for patterns of lexical factors account for patterns of lexical diffusion.diffusion.
The The AoAAoA effect is reliable only for effect is reliable only for low- low-frequency wordsfrequency words
-- Gerhand and Berry (1999)-- Gerhand and Berry (1999)
Gerhand & Berry (1998): Gerhand & Berry (1998): naming latenciesnaming latencies in the in the reading by 33 British college students of 64 reading by 33 British college students of 64 words, divided into the following 4 categories: words, divided into the following 4 categories:
(a) early-acquired, high frequency ((a) early-acquired, high frequency (win, cousinwin, cousin); ); (b) early-acquired, low-frequency ((b) early-acquired, low-frequency (elf, rattleelf, rattle); ); (c) late-acquired, high-frequency ((c) late-acquired, high-frequency (sex, unionsex, union); ); (d) late-acquired, low frequency ((d) late-acquired, low frequency (cue, marvelcue, marvel) )
They found that the two effects , , , were They found that the two effects , , , were ""entirely additiveentirely additive: Participants were faster to : Participants were faster to read aloud early-acquired than late-acquired read aloud early-acquired than late-acquired words and were also faster to name high-words and were also faster to name high-frequency than low-frequency words." frequency than low-frequency words."
Evidence from Adult VocabularyEvidence from Adult Vocabulary
Words for 'heaven', 'priest', and'devil' Words for 'heaven', 'priest', and'devil' pattern with high-frequency words in the pattern with high-frequency words in the Ormulum Ormulum ms., written by a monk.ms., written by a monk.
Words that are frequent within a particular Words that are frequent within a particular occupation under reduction typical of occupation under reduction typical of frequent words, as in frequent words, as in boatswain /boatswain /bosən/, bosən/, coxswaincoxswain /kaksən/, and /kaksən/, and kiln kiln /k/kɪɪl/.l/.
Discourse Influence, as summarized by Discourse Influence, as summarized by Wright (2003):Wright (2003):
"In these studies talkers have been shown "In these studies talkers have been shown to produce to produce more reduced speechmore reduced speech when when contextual informationcontextual information within the within the utterance or in the environment can utterance or in the environment can aid aid the listener in recognising what is saidthe listener in recognising what is said,,
and to produce and to produce more careful speechmore careful speech when when the talker is aware of conditions the talker is aware of conditions that may impede the listener's ability to that may impede the listener's ability to understandunderstand what is said." what is said."
Berg (1998: 243)Berg (1998: 243)
"Frequency speeds up the word-"Frequency speeds up the word-recognition processrecognition process . . . . If speakers . . . . If speakers exploit this principle, they can be more exploit this principle, they can be more sloppy about the pronunciation of high-sloppy about the pronunciation of high-frequency wordsfrequency words than about that of low- than about that of low-frequency items, while still achieving the frequency items, while still achieving the same degree of communicative success. same degree of communicative success. In fact, this is precisely what speakers do: In fact, this is precisely what speakers do: the the higher the redundancy of a word, higher the redundancy of a word, the less accurately it is articulatedthe less accurately it is articulated (Lieberman 1963)." (Lieberman 1963)."
Shortcoming:Shortcoming:
Discourse considerations alone do Discourse considerations alone do NOT account forNOT account for other types of other types of changes, such as changes, such as stress shifts stress shifts oror vowel shifts.vowel shifts.
Misunderstanding #9Misunderstanding #9
That Frequency effects are found over a That Frequency effects are found over a population of speakers but not within population of speakers but not within individual speakers. individual speakers.
“ “Labov’s (1994) assessment of the Labov’s (1994) assessment of the ‘neogrammarian controversy’ provides ‘neogrammarian controversy’ provides striking confirmation of striking confirmation of regular sound regular sound change at the level of the individualchange at the level of the individual, , and and lexical diffusion at the level of lexical diffusion at the level of speech communitiesspeech communities.”.” (Blevins 2004) (Blevins 2004)
Fidelholz (1975):Fidelholz (1975):
Reduced vowelReduced vowel Full vowelFull vowel
astronomyastronomy
mistakemistake
abstainabstain
gastronomygastronomy
mistook mistook abstemiousabstemious
Misunderstanding #10Misunderstanding #10
that that lexical diffusionlexical diffusion is the diffusion of a is the diffusion of a completed sound changecompleted sound change (Blevins 2004) (Blevins 2004) OR OR
"all phonological change starts with lexical diffusion and most ends up Neogrammarian, given enough time" (Lass1997)
Conclusions:Conclusions:Phonological, morphological, Phonological, morphological, semantic, social, pragmatic,semantic, social, pragmatic, and and cognitive factorscognitive factors all influence which all influence which words are affected when.words are affected when.
But But Word FrequencyWord Frequency must also be must also be recognized a key factor in how a sound recognized a key factor in how a sound change spreads through the lexicon. change spreads through the lexicon.
And . . .And . . .
The growing evidence of word The growing evidence of word frequency effects in even very low-frequency effects in even very low-level phonetic output supports the level phonetic output supports the view that what seem to be view that what seem to be Neogrammarian/regular changesNeogrammarian/regular changes must actually be must actually be changes that diffuse changes that diffuse very rapidly over words in the lexiconvery rapidly over words in the lexicon..
