Taalportaal - the digital language portal

Dutch
Frisian
Saterfrisian
Afrikaans

Show all

/t/-deletion before the suffix -st

quickinfo

Stem-final /t/ deletes preceding the suffix -st. For instance (do) praatst /pra:t+st/ (you) talk is realized as [pra:st], not as [*pra:tst]. Tiersma (1979:145-146) was the first one to suggest an analysis of this phenomenon (see also Tiersma (1985:31-31)/Tiersma (1999:28)). His analysis was elaborated in Hoekstra (1985). This has become the commonly accepted approach, also found in, for instance, Popkema (2006:77).

Stem-final /t/ deletes preceding the suffix -st. This is an exceptionless pattern in inflected verb forms, where -st figures as the ending of the second person singular. It is exemplified in (1):

Examples of /t/-deletion before the suffix /st/ in inflected verbs

do switst

/swɪt+st/

[swɪst]

you sweat

do ytst

/it+st/

[ist]

you eat

do ietst

/iət+st/

[iəst]

you ate

do skiftst

/skɪft+st/

[skɪfst]

you sort (out)

do startst

/start+st/

[starst]

you start

do tochtst

/tɔxt+st/

[tɔxst]

you thought

do skeldst

/skɛld+st/

[skɛlst]

you scold, you call names

/t/-deletion occurs in both the present and the past tense and whether or not /t/ is preceded by a consonant, that is, whether or not it is part of a stem-final sequence.

extra

In do tochtst you thought, /t/-deletion results in [tɔxst], which contains the sequence of fricatives [-xs-]. Now, obstruent sequences in general must meet two constraints: preferably, the obstruents do not agree in continuancy, whereas they must agree in voicing (see complex onset: sequences of two obstruents). This means that [-xs-] is a less favoured cluster, though not a prohibited one. It may be repaired by dissimilation, resulting in [-ks-]: [tɔkst].

extra

In do skeldst you scold, you call names, stem-final /d/ ends up in the obstruent cluster /dst/ ( /skɛld+st/). Since obstruent sequences must agree in voicing (see complex onset: sequences of two obstruents) and since word-final obstruents are realized as voiceless (see Final Devoicing), /d/ is adapted to the voicing of the suffix, resulting in /skɛlt+st/. The latter is input to the /t/-deletion at hand.

The phonetic form [skɛlst] is easily interpreted as the verb stem /skɛl/ followed by the suffix /st/, which has given rise to the variant form skelle. The verb with stem-final /d/ is etymologically 'correct', as shown by the fact that both old skelde and new skelle have the preterite stem /skold/ and the past participle /skoldən/, as in hy skold [skolt] he scolded, he called names and hy hat ús útskolden [ytskoldn̩] he has abused us. Confer also the Dutch cognate schelden.

/st/ is also the suffix of the superlative form of adjectives. The examples of superlatives with an adjective ending in /t/ in (2) are taken from Hoekstra (1985):

Examples of /t/-deletion in the superlative of adjectives

wietst

/viət+st/

[viəst]

wettest

geleardst

/ɡəlɪəd+st/

[ɡəlɪəst]

most learned

sêftst

/sɛ:ft+st/

[sɛ:fst]

softest

geskiktst

/ɡəskɪkt+st/

[ɡəskɪkst]

most suitable

frjemdst

/frjɛmd+st/

[frjɛmst]

strangest

korruptst

/kɔrøpt+st/

[kɔrøpst]

most corrupt

lichtst

/lɪxt+st/

[lɪxst]

lightest

bekendst

/bəkɛnd+st/

[bəkɛ̃st]

best known

wyldst

/vild+st/

[vilst]

wildest

Here as well, it does not seem to matter whether stem-final /t/ is preceded by a vowel or a consonant.

There is, however, a difference between the patterns in (1) and (2). Whereas /t/-deletion is obligatory with the verbs in (1), it is less so with the adjectives in (2). The realization of, for example, do ytst /it+st/ you eat is [ist], whereas [itst] is a pure spelling pronunciation and is felt as akward or simply wrong. But although [viəst] is a possible realization of wietst wettest, [viətst] is not ill-formed. The realization of stem-final /t/ is all the more common in case it is part of a consonant sequence; forms like wyldst wildest and bekendst best known can be realized as either [vilst]/ [bəkɛ̃st] or as [viltst]/ [bəkɛntst], no matter how complicated the final sequence of the latter may be.

extra

The stem-final /d/ of geleard /ɡəlɪəd/ learned can only be realized as [t] in geleardst /ɡəlɪəd+st/ most learned, and as such it is subject to deletion: [ɡəlɪəst]. The same holds for the final /d/ of bekend /bəkɛnd/ known in bekendst best known; in this case, /t/-deletion results in the sequence [-ɛns-] ( /bəkɛnst/), so that [ɛ] is subject to Vowel Nasalization: [bəkɛ̃st]. In lichtst /lɪxt+st/ lightest, /t/-deletion results in the fricative sequence [-xs-] ( [lɪxst]), which may be subjected to dissimilation, so that it is realized as [-ks-]: [lɪkst].

Derivational suffixes beginning with /st/ may also trigger /t/-deletion, as exemplified in (3) (examples taken from Hoekstra (1985)):

Examples of /t/-deletion with derivational suffixes beginning with /st/

-ste

hûndertste

/hundət+stə/

[hundə(t)stə]

hundredth

achtste

/axt+stə/

[axtstə/a{x/k}stə]

eighth

miljardste

/mɪljard+stə/

[mɪljartstə/mɪljarstə]

billionth

-ster

arbeidster

/arbajd+stər/

[arbaj(t)str̩]

female worker

hurdrydster

/hødrid+stər/

[høtri(t)str̩]

female speedskater

sprintster

/sprɪnt+stər/

[sprɪntstr̩/sprɪ̃str̩]

female sprinter

-ster

Moddergatster

/modərɡɔt+stər/

[modr̩ɡɔ(t)str̩]

inhabitant of, related to Moddergat

Earnewâldster

/jɛnəvɔ:d+stər/

[jɛnəvɔ:(t)str̩]

inhabitant of, related to Earnewâld

Drachtster

/draxt+stər/

[draxtstr̩/dra{x/k}str̩]

inhabitant of, related to Drachten

Wytgaardster

/vitɡa:d+stər/

[vidɡa:(t)str̩]

inhabitant of, related to Wytgaard

/t/-deletion is less obligatory here than it is in (2). Since place names are locally bound, it may be the case that /t/-deletion is more common for the inhabitants and those living in the vicinity of the place concerned than for those living farther away. Such considerations do not play any role in the case of /t/-deletion in verbs.

Due to /t/-deletion, the sequence /tst/ does not reach the surface. For /t/-deletion to apply, however, it is crucial that /t/ and /st/ belong to different morphemes. The fact of the matter is that the sequence /tst/ as such is not prohibited in Frisian. There are verbs of the first weak class with a stem ending in /ts/, like bats(e) crash down; to wade, bûts(e) to batter, to hammer, to knock, fyts(e) to cycle, keats(e) play a certain ball game, kwits(e) to squeeze; spit out, poets(e) to polish; to brush, and swets(e) to boast, to brag. Inflecting these with -t(e) ( /t(ə)/) yields the obstruent sequence /tst/, which is not subjected to phonological adjustment. The same is found with adjectives ending in /ts/, like bits snappish, snappy, fits sharp, biting, caustic, kwelts lame, lilts angry. When these are inflected with the superlative suffix st, the resulting sequence, /tsst/, undergoes degemination − /tsst/ → /tst/ −, but this does not necessarily pave the way for the reduction of /tst/ to /st/.

extra

An exception is lyts small, the superlative form of which, lytst /lyts+st/, is realized as [list], to the exclusion of [*litst]. The superlative of let /lɛt/ late is lêst /lɛ:st/, in which compensatory lengthening of the vowel /ɛ/ must have followed in the wake of /t/-deletion. These are exceptional forms.

In all likelihood, the phonological motivation for the process at hand was to put a limit to the number of (near-)identical consonants in a tautosyllabic sequence (which may be understood as an OCP-effect). The process, however, has lost this original motivation. It is no longer couched in purely phonological terms, for it has come to be triggered by a specific morpheme, as a result of which it now only applies in the specific morphological context in which this morpheme shows up. This brings with it that the cluster /tst/ can freely occur outside this context.

extra

The (near-)identical consonants are all coronals. It is not without reason therefore that Tiersma (1979:145-146) dubbed his rule Dental Cluster Simplification.

There is another context in which /t/-deletion applies. Subordinating conjunctions are inflected with the suffix -st if the subordinate clause is headed by a verb in the second person singular (in which case the personal pronoun do/dû you (sg.), when not bearing prominence, is generally omitted). In case the conjunction ends in /t/, the latter deletes, see Visser (1988:204-206). This is exemplified in (4):

Examples of the deletion of the final /t/ of subordinating conjunctions in case the subordinate clause has a verb in the second person singular

wy hoopje datst ek meidochst

ik begriep net watst bedoeldest

wy freegje ús ôf, oftst wol wier motivearre bist

/t/-deletion is as obligatory here as with finite verbs − see (1) above −, so realizations like [dɔtst] (datst), [vɔtst] (watst), and [ɔtst] (oftst) are out.

Words belonging to other categories than conjunctions are also used as subordinating conjunctions, in which case they are extended with the 'conjunction morpheme' t ( /t/). Take the demonstrative pronoun dy /di/ that and the adverbs doe /du/ then and wêr /vɛ:r/ where, the various uses of which are illustrated in (5):

The various uses of dy, doe and wêr

as demonstrative pronoun and adverb

ik seach dy man dêr

I saw that man there

I saw that man over there

doe seach ik har

then saw I her

then I saw her

wêr wenje se?

where live they?

where do they live?

as subordinating conjunction

dat is de man [dy't ik dêr seach]

that is the man whom I there saw

that's the man whom I saw over there

[doe't ik har seach] wie it al te let

when I her saw was it already too late

when I saw her, it was too late already

hy wit wol [wêr't se wenje]

he knows all right where they live

he knows where they live

As illustrated in (4), inherent subordinating conjunctions are inflected with the suffix -st if the subordinate clause is headed by a verb in the second person singular. The same happens to 'accidental' conjunctions like those in (5b). The extension with -st makes for a context in which /t/-deletion applies. This is illustrated in (6):

Examples of the deletion of the final /t/ of 'accidental' subordinating conjunctions in case the subordinate clause has a verb in the second person singular

dat is de man [dy'tst dêr seachst]

[doe'tst har seachst] wie it al te let

hy wit wol [wêr'tst wennest]

The forms dy'tst, doe'tst, and wêr'tst have the internal structure /di+t+st/, /du+t+st/, and /vɛ:r+t+st/, respectively, which is also reflected in the spelling. But they are realized as [dist], [dust], and [vɛ:st]. Again, as with finite verbs, /t/-deletion is obligatory here; the realizations [ditst] (dy'tst), [dutst] (doe'tst), and [vɛ:tst] (wêr'tst) are out.

extra

Preceding the conjunction morpheme t, the final /r/ of wêr /vɛ:r/ where is subject to /r/-deletion (see: /r/-deletion in complex words derived with a consonantal suffix).

Taken at face value, the /t/-less forms may also be interpreted such that the suffix -st directly attaches to dy, doe, and wêr, so that the internal structure is /di+st/, /du+st/, and /vɛ:r+st/, respectively. In this connection, it should be noted that people who are less familiar with the Frisian spelling conventions − or not at all, for that matter − often spell these forms as dyst, doest, and wêrst, so without the conjunction morpheme.

This analysis, however, would imply that inflected conjunctions come in two kinds, one for subordinate clauses with a verb in the second person singular (ending in -st) and one for clauses with verbs in the first and third person singular and all plural persons (ending in -t). However, an analysis which only employs the conjunction morpheme -t is to be preferred on conceptual grounds (see also De Haan (1994)). This means that forms like dy'tst, doe'tst, and wêr'tst are best assumed to have the internal structure /di+t+st/, /du+t+st/, and /vɛ:r+t+st/, comprising the conjunction morpheme -t followed by the inflectional suffix -st (2nd.SG).

Arriving at the realizations [dist], [dust], and [vɛ:st] poses no problem, since the process of /t/-deletion must be assumed for independent reasons. It is triggered by the suffix -st, whether the preceding /t/ is part of a verb stem or of a subordinating conjunction.

References

Haan, Germen J. de1994'Inflection and cliticization in Frisian -sto, -ste, st'Nowele2375-90
Hoekstra, Jarich1985t-deletion before suffix-initial st in modern West FrisianNOWELE : North-Western European language evolution563-76
Hoekstra, Jarich1985t-deletion before suffix-initial st in modern West FrisianNOWELE : North-Western European language evolution563-76
Hoekstra, Jarich1985t-deletion before suffix-initial st in modern West FrisianNOWELE : North-Western European language evolution563-76
Popkema, Jan2006Grammatica FriesUtrecht/ LjouwertUitgeverij Het Spectrum BV Prisma Woordenboeken en Taaluitgaven/ Fryske Akademy
Tiersma, Pieter M1979Aspects of the phonology of Frisian based on the language of GrouMeidielingen fan de stúdzjerjochting Frysk oan de Frije Universiteit yn Amsterdam4
Tiersma, Pieter M1979Aspects of the phonology of Frisian based on the language of GrouMeidielingen fan de stúdzjerjochting Frysk oan de Frije Universiteit yn Amsterdam4
Tiersma, Pieter M1985Frisian reference grammarDordrechtForis Publications
Tiersma, Pieter M1999Frisian Reference GrammarAfûk, Ljouwert
Visser, Willem1988In pear klitisearringsferskynsels yn it FryskDyk, dr. S. & Haan, dr. G.J. (eds.)Wurdfoarried en Wurdgrammatika. In bondel leksikale stúdzjesLjouwertFryske Akademy, Ljouwert175-222

⇑ Back to top

About this article

Author(s): Willem Visser

Category: Frisian Phonology

Keywords: t-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