| (In the text I have to use SAMPA
for phonetic transcription, in the figures I use the IPA notation)
In almost all of the Germanic languages and dialects that I know,
the original Westgermanic long monophthongs î and û
have developed into low diphthongs, i.e. /ai/ and /au/. This is
the development you can find in most books about the history of
English and German. The mid diphthongs [Ei] and [9y] are often not
even mentioned. Nevertheless, there is no doubt that there has been
an intermediate stage of [Ei] and [9y] in both languages, even though
this situation did not last very long. [Ei] and [9y] have always
proved to be very unstable.
In the 16th century the /ai/ and /au/ in English en German were
already a fact: wiin had become /wain/ and hoes had
become /haus/. The same is the case in a lot of Dutch dialects.
During the 16th century the dialects which show diphthongization
of î and û, have continued this process
and this resulted in the same diphthongs as in English and German,
/ai/ and /au/ (or /oi/).
These facts are not really very surprising, because lowering of
diphthongs is a very natural development, in languages all over
the world (Labov 1994). Therefore, it is rather remarkable that
the official Dutch pronunciation, the so-called ABN standard, still
has the mid-diphthongs [Ei] and [9y].
There is a Dutch proverb which is ascribed to the famous German
poet Heinrich Heine. It goes as follows: when our planet Earth will
come to an end, better go to the Netherlands because there everything
happens 50 years later than anywhere else. It seems that Heine is
not really the author of this proverb, but in phonology the saying
itself seems to be true, more than true. In ABN-Dutch the Great
Vowel Shift has started not just fifty but more than three hundred
years later than elsewhere, that is, it is taking place now. And
so we can witness a very comprehensive vowel shift taking place
right in front of our eyes (or rather ears), simply by listening
to the Dutch radio and tv stations.
I give some examples. First I discuss a few mid diphthongs of ABN-Dutch.
Concerning the abbreviation ABN, this stands for ‘Algemeen Beschaafd
Nederlands’, in English: General Civilized Dutch. The name is comparable
more or less with RP in English: a pronunciation which gives little
or no clue to a speaker’s regional affiliation (O’Connor 1973, 128).
Sound sample nr. 1 concerns some fragments of ABN-Dutch;
||Click the speaker
to start sound sample nr.1
they are taken from daily radio-programs. The spoken text, written
in normal Dutch spelling, can be followed in figure 1. The
diphthongs in question are printed in bold. Essential fragments are
repeated twice or three times, as the text shows.
|Figure 1. Fragments casual spoken Dutch:ABN
||- Joop en ik die waren dinsdag in Leiden
- .. Leiden bij de overhandiging
- .. Leiden bij de overhandiging
||We luisteren even naar een fragment uitgesproken
op vier mei
- .. vier mei
- .. vier mei
Some indications about the profession of the speakers:
The first speaker is a professor in Dutch linguistics. The second
is a highly competent presentator of tv and radio programs. Both of
them are speaking ABN-Dutch as can be heard from their ij sounds.
Sound sample nr. 2 concerns some fragments of Dutch with lowered
diphthongs, especially with lowered ij.
||Click the speaker
to start sound sample nr.2
I have called this kind of Dutch ‘Poldernederlands’ (Polder-Dutch),
for reasons which I will discuss later. The text of the speech fragments
is shown also in figure 2. About the speakers: the third speaker
is a professor in philosophy; the fourth speaker is a professor in
religious history and the fifth is a presentator from a commercial
|Figure 2. Fragments casual spoken Dutch: Polder-Dutch
||.. vrouw is vaak ik dan jij en bijna nooit zij
- en bijna nooit zij
- en bijna nooit zij
||'t gaat over een conflict tussen de ene groep waar je bij
- ..waar je bij hoort
- ..waar je bij hoort
||- in de rij staat
- in de rij staat
- in de rij staat om auditie te doen voor 'n soap
As can be heard the realization of ij by the last three
speakers comes very near to English and German aai: Compare,
for example, the pronunciation of the Dutch preposition bij
[ba.i] in the fragments you heard last, with English by.
And the vowel in raai from the last speaker is very similar
to that of English right. The differences between those two
phonemes have to do with articulation. The [Ei] requires a more
closed mouth, or a higher jaw. The /ai/ requires a lower jaw.
But also the acoustic features of both phonemes considerable, as
can be seen in figure nr. 3. The pictures show the frequencies
of the two most important formants of the vowels; on the left picture,
that of the formants of the word bij,two black strips in
the lower part, immediately after the b, show that the two
frequencies of ei are not stable but are changing and separating,
during the whole performance of the diphthong. That means that at
any point of the performance the ei sounds the same. Phoneticians
call these type of diphthong an ‘essential’ diphthong; the same
is the case with two other Dutch diphthongs, the [9y]in huis
and [Au] in koud; they have the same kind of features as
The shape of the formants of the aai of baai, which
means ‘baye’, in the right picture, is totally different.
Here we see after the b that the two strips are parallel,
during the performance of the aa; they separate out only
after the aa, when the half-vowel j begins. So it
is clear that the aai in baai is a completely different
sound from the [Ei]; in fact it is a combination of aa and
j, a sequence of two vowels. When we drop the last part,
there is no diphthong at all. The lowered [Ei] in Polder-Dutch is
supposed to have the same type of features as this aai.
There are two main causes, in my opinion, which can explain why
the mid diphthongs in many languages are lowered very quickly. Both
causes are of an internal type. First of all, consider the situation
from the point of the speaker. As phoneticians have pointed out,
low diphthongs are easier to articulate because they don’t need
a very accurate articulation: you open your mouth and the aai
and aau are there, so to speak. But not such with the ABN-diphthongs.
For non-native speakers, who are learning to speak Dutch, they are
the most difficult Dutch vowels. Normally they are the last traces,
from which you may detect that a speaker is a non-native one.
From the point of view of the listener, there is also an internal
reason to prefer low diphthongs: a low diphthong is louder, because
the mouth is wide open, so you can hear it better. It is because
of that I think that the most lowered diphthongs are produced by
singers of Dutch songs. Both factors, which favour lowered diphthongs,
make it highly remarkable that ABN has preserved its mid diphthongs
for such a long time. There must have been a very strong factor
behind this. And there was.
In the same period that the vowel shift took place in English and
German, and also in the dialects of the Dutch provinces of North
and South Holland (the 16th century) there was a movement in the
Netherlands to create and establish a language which could function
as a nationwide language for the time when the Dutch Republic should
become independent. A number of writers, linguists, teachers and
other prominents were engaged in selecting and cultivating elements
for the national language (the later ABN). Those elements were mainly
taken from the dialects and the language of educated people in the
provinces of Holland. With respect to the diphthongs, there was
an almost general opinion that the diphthongs as they were spoken
in the rural dialects of Holland, especially aai, were not
acceptable: they sounded too awful. Therefore all teachers, writers
and other important citizens decided to stick to [Ei] and [9y].
So strong was the influence of the teachers and so great the prestige
of the language they had created (or the respect for the people
who spoke that language), that the mid diphthongs, ei, ui
and ou, could survive and almost remained unchanged for centuries.
This situation continued until the sixties and seventies of this
century. In this period society changed in a radical way, as was
the case in most of the western societies. Rules and traditions
were falling away. The stronger the old rules the stronger the reaction
Also the ABN standard, which is - as I said earlier - the pronunciation
of Dutch which was taught at school and was the spoken standard,
began to lose its prestige and functioned no longer as a model.
So at last it became possible for the three essential diphthongs
to be lowered. Not the lowering itself is surprising, because it
is a natural change, but it is that long delay of more than three
centuries that is remarkable.
|Figure 4. Shift of long vowels and diphthongs in ABN-Dutch
The lowering of the ABN diphthongs is very similar to the Great
Vowel Shift in English, as can be seen in figure 4. Like
the G.V.S., it is a complex of different chain-movements. In the
left part of the vowel-triangle, that is the front part of the human
mouth the diphthong /e/ is
being lowered and reaching the position of /ai/. The place of the
original[Ei] is being filled up by the lowering and diphthongization
of the long monophthong /e./.
The same happens with the two others diphthongs, [9y] and [Au].
Lowering of the [9y] makes room for the [ø] to be lowered
in turn; lowering of the [Au], gives the opportunity for [o.] to
do the same. When you are listening to Dutch speech the change of
the [Ei] is especially striking, more than that of the two other
diphthongs. There are two reasons for this. First of all, the diphthong
[Ei] is much more frequent than the other two. Secondly, the change
of [Ei] is more radical, with regard to its acoustic properties.
There are very clear indications that the change of the diphthongs
has taken place earlier and has been more drastic than that of the
long monophthongs above them. It, therefore, seems correct to assume
that the change is a drag-change, but further research has
to be done on this point. And again, not the change itself is exceptional,
but rather the extremely long delay.
There is yet another surprising aspect. It is a well-known fact
in sociolinguistics that in most situations women speak a variety
of language that is nearer to the standard than men do. Women are
more aware of the social significance of linguistic variables than
In the Netherlands this situation obtained too, as is pointed out
in several investigations. Now, however, a complete different development
seems to be in progress. The first group of speakers for which the
lowered diphthongs were noticed, were high middle class women aged
between thirty and forty. These are writers, scientists, politicians,
film directors, professors and so on. They live all over the Netherlands.
They are ambitious and intend to make a career or have already made
one. A number of them have declared themselfs feminists. No men
in a comparable group have been observed to have the same lowered
For the first time then we are confronted with a language change
which is moving away from the ABN standard, and which started with
well-educated women in the seventies. Since Polder-Dutch is spreading
among the whole population, step by step, the younger the generation
the more they speak Polder-Dutch. Among children younger than ten,
male and female, the lowered diphthongs are the norm. Parents who
speak ABN themselves are very concerned when they hear their children
speak in such a way. But all their efforts to change their childrens
aai’s in ei’s are useless. For those parents this
is painful because in the Netherlands these lowered Dutch sounds
has the same effect on ABN-speakers as you perhaps would experience
if you heard Queen Elisabeth speak Cockney English, for example,
while in Heathrow asking for gaait number foif (‘Gate number
Of course there is a lot of investigation to do, because a number
of aspects are involved, such as the history of the feminist movement
in the Netherlands. Also the phonetic and phonological aspects have
to be considered. Perhaps there is also a kind of lexical diffusion.
An investigation in real time will be set up. That seems possible
because many of these women have spoken on radio in the past and
their speeches and interviews have been taped and been preserved.
Also a classical sociolinguistic investigation will be carried out,
with recordings of different styles of speech, casual, formal, word
lists and so on, to find out if this really is a change in progress.
It would also be very interesting to investigate whether the lowering
of the diphthongs is a linguistic change from above, which I believe,
or from below.
The first small steps have already been taken and I am in a position
to show some results. They represent our first attempt to try to
confirm our observations that well-educated women use lower diphthongs
than educated men do. We made recordings from a well-known sophisticated
tv-program on the 3d channel, which is called The Blue Light.
In this program four people critically discuss recent tv-programs.
Two members of the panel change every time, the so-called guests.
In last year’s season it so happened that each time these guests
were a man and a woman, all of higher middle class.
One of the students of Dr. Renée van Bezooijen, from the
University of Nijmegen, Mrs. Edelman, has transcribed the first
15 realizations of the three diphthongs of each of the in total
12 guests. This number is supplemented with, as far as the first
diphthong is concerned, 3 men and 3 women from a radio news program.
So in total we tested 18 speakers: 9 women and 9 men.
|Figure 5. Mean Scores
|Guests TV-program 'Het blauwe
|Adriaan van Dis
|Govert van Brakel
The realizations are indexed in 4 steps, as you can see at the
bottom of figure number 5. When a diphthong has the same
quality as in ABN, it receives a 0, the most lowered aai
receives a 3; the two steps in-between are indexed as 1 and 2. After
the name of each speaker his or her mean scores per diphthong are
presented; in column number 4 the mean of their three diphthongs.
The higher the score the lower the diphthong. From a first glance
at column 4 one may conclude that in most cases (but not always)
in each pair, the woman has the highest scores, which means the
Remarkable are the scores for Roberta Alexander. She is a famous
American classical singer, but she has lived in the Netherlands
for years. She speaks Dutch very well, although with an unmistakable
American accent. Nevertheless her diphthongs are not as low as those
of two Dutch native-speakers, Dana Nechustan and Annemarie Mol.
On this point Roberta Alexander speaks more ABN-like than the two
Dutch women. But because she is not a native speaker we have left
her scores aside.
To test the significance of the differences in the mean values
for the three variables and for the total values we used a t-test
for paired samples. For [Ei] the test was based on 9 paired values,
for [9y], [Au] and /total/ on 6 paired values. The results have
been presented in figure nr. 6.A 1-tailed test was used because
our hypothesis was one-directional: we hypothesized that the female
speakers would have higher values than the male speakers. It can
be seen that the values for [Ei] and /total/ are significant at
the 5% level, for [9y] even at 3% level, with the means in the expected
direction: we indeed find higher values, i.e. lower realizations,
for the female speakers.
So this means that we can maintain our hypothesis. Of course, we
will continue our investigations with greater samples and samples
of a different kind. But for the moment we may conclude that higher
middle class women use lower diphthongs than men from the same class.
In other words, Netherlandic higher middle class women take the
lead in speaking a variation of Dutch which is moving away from
ABN-Dutch. This is in full contrast with the situation in previous
generations. There we saw that lower class men were moving away
from the standard norm. So the development of modern Dutch is in
double contrast with that of the older situation: women of the higher
middle class are now playing the same role as men of the lower class
in the past.
But why is it the only women from this class do this? Of course
it is linked with the position of women in our society. Women, who
in the past showed awareness of the social significance of linguistic
variables, now are the first to see that in our egalitarian society
language functions no longer as such. Besides, the women who are
using these lowered diphthongs, have good jobs and important positions;
they profit the most from the new economic and social situation.
And therefore they don’t need language to acquire status.
This can explain their linguistic behaviour. The moment that rules
and traditions, including those concerning pronunciation, are undermined
or thrown away, the groups which have respected those rules the
most, are inclined to react in a more extreme way than others. Generally,
the greater the repression the more extreme the reaction, when the
repression has gone. A second important factor is the feminist attitude
among a number of those women, which is responsible for their increasing
I have called this lowered variety of Dutch Poldernederlands,
which in translation is something like Polder-Dutch, for the following
reason. ‘Polder’ is loaned from the composition ‘poldermodel’, a
popular name for the recent economic model of the Netherlands, a
model that is based on consensus of all parties. Everyone knows
that this poldermodel has brought a great deal of welfare to our
country. But in the meantime this economic success is changing our
society in a less positive way in some respects. It has also strengthened
the social results of the sixties and seventies. One of these results
is Poldernederlands. As you understand it is also a kind of tongue-in-cheek-name.
Finally we may conclude that Heinrich Heine or his ghostwriter
has been only partly right: in the Netherlands the lowering of the
mid diphthongs indeed happens very much later than anywhere else,
but on the other hand Dutch society has token a lead in becoming
an advanced egalitarian society, which allows well-educated women
of the higher middle class to say ‘Farewell to linguistic variables-with-former-prestige’.
Gender variation in Dutch. A sociolinguistic study of Amsterdam
speech, Dordrecht 1989.
Principles of Linguistic Change. Volume 1: Internal Factors,
Phonetics, Middlesex 1973
Article published on the WWW: October 1999