Taalportaal - the digital language portal

Dutch
Frisian
Saterfrisian
Afrikaans

Show all

/h/-deletion

quickinfo

A restriction on the occurrence of the voiceless laryngeal fricative /h/ is 1) that it can only occupy the syllable onset and 2) that it cannot be part of a consonant cluster. In other words, /h/ must immediately precede a vowel (see the glottal fricative /h/). The outcome of phonological processes may be that /h/ ends up preceding a non-vowel. If, for instance, a stem with a centring diphthong began with /h/, the result of Breaking (see Breaking: the environment) was the sequence /h{j/w}/. Since this configuration is ill-formed − /h/ does not immediately precede a vowel here − it was in need of repair, which was implemented by the deletion of /h/. This is shown by alternations like hier /hiər/ hair ~ hierrich [jɪrəx] hairy and hoas /hoəz/ stocking ~ hoazzen [vwa:zn̩] stockings. /h/-deletion is the subject of this topic.

Due to the orthographic principles of uniformity ('the spelling of a morpheme should be kept constant') and etymology ('the spelling of a word should take the etymology of that word into consideration'), more <h>s are written than are pronounced. Take the word hjouwer oat(s). Its pronunciation is [jɔwər], which is indistinguishable from that of jouwer /jɔw+ər/ giver; it is, however, written with <h>-, which can be motivated on etymological grounds, see the cognates haver (Dutch) and Hafer (German). Take, furthermore, an alternation such as hier /hiər/ hair ~ hierrich [jɪrəx] hairy; though hierrich is pronounced without [h-], it is written with <h>-, because of its relation with hier.

An overview of the contexts in which word-initial /h/ has been deleted is given in (1).

The contexts in which word-initial /h/ has been deleted

Words in which /hj/- did not result from Breaking

hjel#dei

/jɛl/

Christian holy day

(related to Old Frisian

hēlich

holy

)

hjoed

/juəd/

today

(cf. Dutch

heden

nowadays

huidig

present, current

, and German

heute

today

)

hjouwe

/jɔwə/

kind of fishing net

(cf. Dutch

haaf

kind of fishing net

)

hjouwer

/jɔwər/

oat(s)

(cf. Dutch

haver

and German

Hafer

)

hja

/ja/

she; they

hju

/jø/

she; they

(related to Old Frisian

hia

and

hio/hiu

)

Examples of words with a broken diphthong without a synchronic counterpart with a centring diphthong

heakkel

/jɛkəl/

rake for cleaning plants out of ditches

hjerring

/jɛrɪŋ/

herring

(cf. Dutch

haring

)

hjerst

/jɛst/

autumn

(cf. Dutch

herfst

)

hjidde

/jɪdə/

flax waste

(cf. Dutch

hede

hee

)

hjir

/jɪr/

here

(cf. Dutch

hier

)

hoanne

/vwanə/

cock

(cf. Dutch

haan

)

hoarrel

/vwarəl/

whirlwind

(cf. English

hurl

)

hoart

/vwat/

while, time; thrust, dig

(cf. Dutch

hort

)

hoastje

/vwasjə/

cough

(cf. Dutch

hoesten

)

Examples of synchronic alternation between a centring and a broken diphthong

hier

/hiər/

hair

hierrich

[jɪrəx]

hairy

heak

/hɪək/

hook

heakje

[jɛkjə]

hook (up)

heal

/hɪəl/

half

heal#wei

[jɛlvi]

halfway

heap

/hɪəp/

heap, pile

heapje

[jɛpjə]

heap, pile (up)

hoed

/huəd/

hat

huodden

[vwodn̩]

hats

hoer

/huər/

whore

huorkje

[vworkjə]

visit prostitutes

hoet

/huət/

giddy-up

huottelje

[vwotl̩jə]

shake, swing

hoale

/hoələ/

cave, hole

hoalling

[vwalɪŋ]

fingerstall

hoas

/hoəz/

stocking

hoazzen

[vwa:zn̩]

stockings

Examples of dialectal alternation between a centring and a broken diphthong

hear(e)

/hɪər/

hearr(e)

/jɛr/

to hear; listen to

hoerde

/huədə/

huorde

/vwodə/

hoard

As is clear, /h/ has systematically been deleted in case it preceded a glide. From a synchronic view point, Breaking is an opaque phenomenon, so it cannot be dealt with in purely phonological terms. The relation between a form with a centring diphthong and the one with the corresponding glide + vowel sequence therefore is one of allomorphy. For cases such as those in (1c-d) two stem allomorphs must be assumed, one with /h/ followed by a centralizing diphthong and one with merely a rising diphthong. The words in (1a-b) can safely be assumed to no longer have /h/ in their underlying representation.

extra

The following words used to be spelled with initial <h>: hwa who, hwat what, hwannear when, hwant because, for, hwer where. Since <h> was no longer pronounced, it was removed from the spelling due to the 1976 spelling-reform. The current spelling therefore is: <wa>, <wat>, <wannear>, <want>, <wêr>, which is more in line with Dutch orthography. The former role of /h/ is still manifested by the English spelling.

extra

The adjective hjit /jɪt/ hot and the verb hjitt(e) /jɪt/ be called, be named have the variants hyt /hit/ and hit(e) /hit/, whose /i/ must have developed out of the centralizing diphthong /iə/.

extra

Though the ill-formed sequence /h{j/w}/ has in general been repaired by the deletion of /h/, in principle this could also have been implemented through deletion of the glide. There are a few instances of the latter. The underlying representation of the noun heakkel rake for cleaning plants out of ditches is either /jɛkəl/ or /hɛkəl/, both of which are a continuation of /hjɛkəl/. And the collocation it hoe en wat the ins and outs (literally: the how and what) has the variant it hoe en het, in which het /hɛt/ has developed from hwet /hwɛt/ what. Deletion of /h/, however, was the unmarked case. This may have to do with the fact that /h/, having no supralaryngeal features, is the weakest consonant, just like schwa is the weakest vowel. Due to its minimal phonological specification, /h/ is prone to deletion, which is a minimal step.

Not only was /h/ deleted in case it preceded the broken diphthongs /wa/ and /wo/, it was also replaced by the labiodental fricative /v/, as shown by words like hoanne [vwanə] cock, hoart [vwat] thrust, dig; time, while, huorde [vwodə] hoard, hoazzen [vwa:zn̩] stockings, and huodden [vwodn̩] hats. This appearance of /v/, however, is an independent phenomenon, which also manifested itself in words beginning with just a broken diphthong, like oarberje [vwarbərjə] consume, dispatch, oarder [vwadr̩] order, oargel [vwarɣəl] (pipe) organ, oarre#mem [vwarə-] grandmother, Oark [vwark] men's name, oartsen [vwatsn̩] old coin (farthing), and uorre#moarn [vworə-] the day after tomorrow. This means that words with initial /h/ + broken diphthong and those beginning with just a broken diphthong have acquired the same onset as those beginning with the sequence /v/ + broken diphthong, examples of which are woarst [vwast] sausage, woartel [vwatl̩] root; carrot, and wuolje [vwoljə] to wind; to wrap (up). All in all then it is not the case that initial /h/ was 'replaced by' the labiodental fricative /v/. All one can say is that the deletion of the former paved the way for the development of the latter.

There is an independent piece of evidence for the presence of /v/ here. In the Wâldfrysk dialect, the sequence labial consonant + /w{a/o}/ has generally turned into labial consonant + /j{a/o}/, which means that the glide /w/ has been replaced by the glide /j/ (see replacement of the glide /w/ of the broken diphthong /w{a/o}/ by /j/ following labial consonants). This has also affected words like those above, so hoanne [vwanə] cock, oargel [vwarɣəl] (pipe) organ, woarst [vwast] sausage, and huodden [vwodn̩] hats, for instance, are now realized with initial [vj{a/o}], as [vjanə], [vjarɣəl], [vjast], and [vjodn̩], respectively. It is due to their initial labiodental /v/ that the words in question could, and did, join this pattern.

The glide /w/ then seems to be prohibited in word-initial position. Though segments with the lowest degree of sonority are favoured in the onset position, glides are not prohibited there, as shown by the large number of words which begin with /j/. There is, however, a difference between /j/ and /w/. Diachronically speaking, the word-initial bilabial glide /w/ has turned into the labiodental fricative /v/, as a consequence of which the newly arisen word-initial /w/ was likely to be felt as deviant. The glides are realized with some (phonetic) frication in word-initial position. This was easily reinterpreted as a consonantal feature, the more so because of /w/'s exceptionality there. The fact that /w/ was reinterpreted as /vw/ need not come as a surprise: on the one hand, labio-dental /v/ and labio-velar /w/ have much in common whereas, on the other hand, they are sufficiently different to count as separate segments.

extra

The pronunciation of the initial part of words like hoazzen stockings, woarst sausage and hoanne cock, rooster has puzzled Frisian phonologists. Is it [vw-], [w-], or something in between? Van Coetsem (1951:90) perceives a kind of more or less lengthened bilabial initial w, or perhaps more properly: a w with a more or less vocalic 'grace note' uw [translated from Dutch] (see also Van Coetsem (1952) and Van Coetsem (1953)). This is taken up in Cohen (1959:125), where the sound(s) in question is (are) transcribed as [w:] or [uw], which is assigned the phonological value /ww/, though it is noted that all this merits further phonetic inquiry. Van Coetsem assumes lenghtened w or uw to be vw phonologically, a view adopted in the second edition of Cohen et al. (from 1961), where (on page 135) the word woarst is transcribed as [ʋu̯ast].

In the literature on Frisian phonology, the labio-dental fricative /v/ on the one and the bilabial glide /w/ on the other hand have been analyzed as realizations of the phoneme /w/, see Fokkema (1938:37) and Cohen (1959:119-120,125). This lead to the confusing proposal to represent [vw-] as /ww-/ phonologically: woartel /wwatəl/. This representation constitutes an OCP-violation, so it can be excluded on principled grounds. A sharp distinction should be made between /v/ and /w/ (see also Visser (1997:63-65)).

References

Coetsem, Frans van1951Over en naar aanleiding van een Nieuwfriese AnlautwijzigingHandelingen der Zuidnederlandse Maatschappij voor Taal- en Letterkunde584-93
Coetsem, Frans van1952Oudfries hw- > w- en Nieuwfries hoâ- > woâ-Leuvense bijdragen, Tijdschrift voor Moderne Philologie4253-57
Coetsem, Frans van1953Nogmaals Fries hoanne-woanneLeuvense Bijdragen, Tijdschrift voor Moderne Philologie4391-101
Cohen, Antonie, Ebeling, C.L., Eringa, P., Fokkema, K. & Holk, A.G.F. van1959Fonologie van het Nederlands en het Fries: Inleiding tot de moderne klankleerMartinus Nijhoff
Cohen, Antonie, Ebeling, C.L., Eringa, P., Fokkema, K. & Holk, A.G.F. van1959Fonologie van het Nederlands en het Fries: Inleiding tot de moderne klankleerMartinus Nijhoff
Fokkema, Klaas1938Inkelde opmerkingen oer it Fryske KlanksysteemFrysk Jierboek36-45
Visser, Willem1997The Syllable in FrisianVrije Universiteit AmsterdamThesis

⇑ Back to top

About this article

Author(s): Willem Visser

Category: Frisian Phonology

Keywords: /h/-deletion

Version: 1.0, December 2015

Version history:

version	editor(s)	date	remarks
1.0	Willem Visser	December 2015

printreport errorcite

Jump to:

introduction
references
related articles