Akan Teleteaching Course 


6.4: Notes on pronunciation

Downstep chat 1: Downstep and Downdrift

Student 1:
(to Ghanaian friend) The other day, I heard the teacher say something about 'downstep'. Can you tell me more about it?
Ghanaian friend:
I have not heard of it. I do not think we do that in Akan. I never did it, none of my family does it. Here comes the teacher, let's ask him.
Teacher:
(laughs) Of course, my friend, you have downstep. The Akan language would not be Akan without the downstep. And you practice it a thousand times every day. One could almost say: Downstep is the most important sound of the Akan language!
Ghanaian friend:
Eeh? Ampa?
Teacher:
Oh yes! You know the proverb which says that people are different?
Ghanaian friend:
Of course. We say: EÀtiárä nyìnaáaá nàsñáThat means: "Peoples' heads do not resemble each other."
Teacher:
(to students)Moá!Listen to the melody.
Ghanaian friend:
Student 2:
It goes pretty much up and down! It reminds me of what we said the other day about tone: You have to learn for each syllable - sometimes even less than a syllable - whether it is pronounced High or Low.
Teacher:
Listen again and try to hum the melody!
Ghanaian friend:
Student 3:
(hums the melody) Well, the whole proverb begins High and ends Low.
Student 2:
But look at the way it is written. There are three words. If you listen carefully you notice that each word begins Low and end High. (Repeats one word after the other.)
Student 1:
So who is right? Both cannot be right.
Teacher:
Yes, both are right. Listen to the playback: E-ti-ri nyi-naa n-sñ[Slow recording] Try to listen to the higher tones of each word and compare them: …tiri …. naa … sñ
Student 3:
It is like going down the steps of a staircase.
Teacher:
Exactly. The first thing to note if you are going to understand the concept of Downstep is this downwards movement of the sentence melody. In a normal statement it goes down step by step. Now concentrate on the lower tones at the beginning of each word and compare them: E … nyi … n … (Slow recording)
Student 2:
The same. Like going downstairs - step by step.
Student 1:
Is this what you call 'downstep'?
Teacher:
Gently! It is not totally wrong to call it 'downstep'. Some linguists (e.g. Stewart (1983b), Tufuor (1982)) do indeed call it 'automatic downstep'. But this is a bit confusing because the Downstep for which particularly West African languages are famous is not automatic but must be learnt.
Student 2:
What do you mean by 'automatic'? It seems to me that nothing happens automatically in Akan, you have to learn by hard work a lot of things which you would not even dream of in learning a European foreign language.
Teacher:
'Automatic' just means predictable. Automatic features are always there if certain conditions are fulfilled. Given these conditions, you can predict that the feature will be there. For instance, in English and German, the phonemes p, t, k are always aspirated at the beginning of a word. The rule is: no choice, no contrast, no meaning.
Student 2:
Are you saying that there is a rule which causes each succeeding High tone to be lower than the preceding High tone? E-ti-ri nyi-naa n-sñ.
Student 3:
And each succeeding Low to be lower than the preceding Low!
Teacher:
Correct, as far as this example goes. This is what some have called automatic downstep, in contrast to non-automatic downstep. But we will call it 'downdrift', and what they call non-automatic downstep we will call downstep. It is important to keep downdrift and downstep apart!
Student 2:
So what is 'Downstep'?
Teacher:
I shall explain in a moment. But first, listen to the following sentence.
Ghanaian friend:
(reads) Paápaá näeá "It is goodness."
Teacher:
(to student 3) Will you repeat?
Student 3:
Ghanaian friend:
(laughs) What you just said means: "It is the 'papa' festival." Paápaá näeáThat means "It is goodness." Paápaá niáeá means "It is the 'papa' festival."
Teacher:
Both sentences have only High tones …
Student 2:
Now I am totally confused. We just said that High tones get progressively lower. And now we have two sentences…
Teacher:
To make the confusion complete, remember that we also had paàpaà 'fan'.
Student 2:
This is Low - there is no confusion.
Teacher:
(to Ghanaian friend) Will you read these three:

 
That means "It is our goodness."
It means "It is our 'papa' festival."
This means "It is our fan."

Ghanaian friend:
(reads)
Teacher:
Notice that papa'fan' is different when it comes after yñàn.
Student 3:
I heard it High-Low.
Teacher:
Moá!! (= well done! -> Unit 8)
Student 2:
But before, when we had paàpaà, it was Low-Low. Why does it change?
Teacher:
Oh, do you not remember, we've already tackled this in Unit 4. (-> Link tone). Listen again to the melody of the three papa. They all begin on High tone. Just concentrate on the the second syllable. First compare its tonal height to the tonal height of the preceding syllable pa…
Ghanaian friend:
(reads) yñn papa nie"its our goodness" --- yñn papa nie"its our festival" --- yñn papa nie"its our fan"
Student 3:
In 'goodness', the second pais the same as the first: paápaá. In the two other words, the second pais lower than the first.
Student 2:
I even hear the third still lower than the second. It goes down very much: paápaà.
Student 3:
To be honest, I heard that too. But I did not trust my ears because I remembered that we had said that Akan distinguishes only two tones, High and Low.
Teacher:
Okay, we shall see that. Let's first compare the tone of the second pato the tone of the following item niewhich means "It is …". yñn papa nie - yñn papa nie - yñn papa nie…
Student 2:
Are they really all different?
Student 3:
Compared to nie, the syllable pawhich precedes it seems pretty much the same height in the first two words, but it is clearly lower in the third word.
Teacher:
Let's try sol-fa notation:

 
e


 
1. yñànà
paá
paá
näeá
2. yñànà
paá
ápaã
ánïãeã
3. yñànà
paá
ápaà
nïãeã

The last papais High-Low as we said, the first has High on both syllables, it does not go down at all. And the second - well I hear it in between.

Student 2:
Can I hear it again?
Teacher:
Of course. Playback. You can even see the difference.
Student 1:
That helps. In the 'goodness'-papathe second pais just as high as the first pain the 'festival'- papa is clearly a bit lower but not as low as in the 'fan'-papa.
Student 3:
So if we have three different tone heights which serve to distinguish words as do the tones on the second syllable of papa- why do we then not say that Akan has three tones - High, Mid and Low? After all Chinese also…
Student 4:
…has four tones. My friend who studies Chinese always said that a tone language worth its name…
Teacher:
Let's leave Chinese aside for the moment. Indeed it is not entirely wrong to think that Akan has three tones. This is what Christaller thought, for instance.
Student 1:
Who is Christaller?
Teacher:
Johann Gottlieb Christaller is the author of the famous Twi-English dictionary. He also published the first really scientific grammar of Akan (which he called Twi) already in 1875. And he also wrote a small booklet on the tones (1893). He was the first to write the tones, and the first to understand how tone works in Akan. As a matter of fact, he was the first to give a scientifically accurate description of an African tone language in general. In both the Grammar and the tone booklet we can see that he had understood that what he called the Mid tone behaved differently from the other two tones. It is essentially for this same reason why today we prefer not to speak of a Mid tone but of a downstepped High, in short Downstep.
Student 2:
That is nice. But I am totally confused because a moment ago we had said that succeeding High tones get lower as the sentence progresses. And now we have this first papaexample where you must stay High if you do not want to be misunderstood.
Teacher:
Yes, let me give you another example.
Ghanaian friend:
Student 3:
The melody seems to remain High. There is no - how did you call that again …?
Student 2:
No, downdrift. But at the end it goes down.
Teacher:
Now you are very close to the solution!
Student 2:
Well, I guess if there are no Low tones but only High, there is no downdrift.
Student 1:
And if there are only Low and no High tones? Or maybe that does not even happen.
Teacher:
It does, and quite often. Listen!
Ghanaian friend:
Ïkõõ Nkran"(S)he went to Accra."

Yñhunuu Ado"We saw Ado."
Student 1:
(tries) Ïkõáõà Nkran.
Ghanaian friend:
Student 1:
Is that not what I said?
Teacher:
What you said was: "When he or she went to Accra …"
Student 3:
So it is as I thought a moment ago: If one Low follows another Low, both stay at the same level.
Student 2:
And if one High follows another High, both stay High.
Teacher:
As we said. But did you notice that at the end - generally on the last syllable - it goes down?
Student 2:
I noticed it. I did not dare to ask.
Teacher:
Student 2:
I thought it is just like in my language, the melody goes down when you come to the end of the sentence.
Teacher:
This is the way it should be and this is the way it is. Listen to the following:

Mñkõ hõ [Mñákõá hõã]
Mñkõ hõ Wukuada [Mñákõá hõá Wuàkuáaádaá] 

How do you hear the melody of Mñkõ hõin the two short sentences?

Student 3:
High High and then it goes down on That is the first. Now, the second. Here does not seem to go down. I think is just the same height as mñkõ…
Teacher:
In the first, is lower. Why?
Student 1:
I guess when it occurs at the end of the sentence, it gets lower.
Teacher:
Yes, we have a sort of Downdrift which always occurs at the end of the sentence, mainly on the last syllable. The last syllable is lowered by just one step before pause.
Student 1:
And if it is a Low tone, can it still get lower.
Teacher:
Yes, listen!
Student 3:
So this means, we have Downdrift on every last syllable?
Teacher:
Yes.
Student 2:
So what about Downdrift in the middle of a sentence?
Student 3:
I think it only occurs when a High follows a Low, or a Low follows a High.
Student 2:
That is if there is a change in tone?
Teacher:
Quite. Let's look at our first example again:

 
eàtärä
nyìnaáaá
nà-
sñá
heads
all
NEG
resemble
"People are different." (Proverb)

-> spectrogram with analysis

Student 1:
Now it strikes the eye. This way it is easy to understand even if you are not a professional linguist. Could you not have shown that at the beginning? It sounded terribly complicated to me. Now I think it is quite simple: Each time the tones change, whether from High to Low or from Low to High does not matter, the tone following the change is one step lower that than the closest preceding tone of the same kind. If the tone following the change is High, this High is one step lower than the closest preceding High. If it is Low, it is one step lower than the closest preceding Low.
Downdrift principle: In a sequence of a High tone followed by one or more Low tones which is in turn followed by one or more High tones, the subsequent High tone(s) are pronounced one step lower than the preceding High tone, due to the lowering effect of the intervening Low tone. In a sequence of a Low tone followed by one or more High tones which is in turn followed by one or more Low tones, the subsequent Low tone(s) are pronounced one step lower than the preceding Low tone, due to the lowering effect of the intervening High tone.
Teacher:
It is true, I could have explained this at the beginning. But I think we have understood a lot in the discussion, and we have maybe understood it so that it will stick.
Student 4:
I hope so. It is always simpler once it has been explained.
Student 2:
I am still confused about the papa.
Teacher:
Yes, we must now talk about 'real' downstep. Downstep is what explains the difference between the first and the second papaDownstep accounts for many important differences in the vocabulary and even more so in the grammar. Therefore we must take the trouble to understand what happens. But if you understand Downdrift you will also understood Downstep because Downstep is so to speak a by-product of the Downdrift principle.
Student 2:
Still sounds mysterious to me.
Student 4:
In Chinese …
Teacher:
(sharply) There is no downstep in Chinese. Downstep really has been invented in Africa and for Africa!
Student 4:
But I have read that in English, too, …
Teacher:
(more sympathetically) True. The concept of downstep comes from Africa. But just as the vowel harmony parameter of ATR (= Advanced Tongue Root ), which we had seen the other day, it has been exported and has been found very useful for describing the intonation of English and other languages, too. Let's keep all this for a special session for those who are interested.
Student 1:
Cannot you give us a simple rule - that is all we can swallow. And then we shall do our best.
Teacher:
Let me give you an example, and you will find out the rule for yourself.
Student 2:
Okay, we will try.
Ghanaian friend:
Fa me kyñ no ma me"Give me my hat!"
Teacher:
Listen to the first part only: Fa me kyñ no… "Take my hat DEFINITE-ARTICLE …"
Ghanaian friend:
Student 4:
(whistles the melody)Fais Low (-> Imperative), meis High, kyñis Low.
Student 2:
But not as low as the preceding Low of the fa! But following the Downdrift principle, if kyñis Low it should even be lower than fa!(to Ghanaian friend) Please, can you say it again?
Ghanaian friend:
Student 4:
So it is Downdrift?
Student 2:
But as you said yourself meis High, and if kyñis not Low because it is higher than fawhich is Low, then kyñmust be High. If it is High and is lower than the preceding High, then there should be a Low between the two High's which causes the second High to be lowered one step.
Student 3:
Yes, and I do not hear a Low in between. So the rule we just established after all this discussion is again wrong. I am now tired …
Teacher:
Give me one more chance. You know what kyñmeans in this sentence?
Student 3:
It means 'hat'.
Teacher:
Right. But now, if I asked you: Tell me the word for 'hat' - what would you say?
Student 3:
I think it is something like ñkyñ.
Ghanaian friend:
Yes, ñkyñ; ñkyñ nie"It is a hat."
Teacher:
So what are the tones of ñkyñ.
Student 3:
I think first Low, then High.
Teacher:
Yes, remember the nominal prefix is normally Low. So, we do have a Low occurring before kyñ.
Student 2:
But it is not there in the example you gave - what was it again?
Ghanaian friend:
Student 2:
I do not hear ñkyñ- just kyñ! There is no Low tone between the meand the kyñwhich would cause the High of kyñto become lower.
Teacher:
Yes, there is.
Student 2:
But it is not pronounced!
Teacher:
True. But it is there.
Student 2:
What does that mean: It is not pronounced but it is there?
Student 3:
Well, I guess it must be there in the Akan speaker's mind.
Student 4:
Sounds weird. Something like a phantom tone.
Teacher:
Some have called it like that. For instance in German: 'Phantomtoene'. In English we have settled for 'floating tone', that is a tone which is there - as you said in the speaker's mind - but is not associated with a word.
Student 4:
How come that a tone can so to speak fall from the waggon?
Teacher:
That is well put. Christaller found out more than a hundred years ago that this happens indeed to tones.
Student 4:
But how?
Teacher:
Remember the word taken by itself is ñkyñIt has two syllables, the first - what we call the prefix - consists only of a vowel, the second - the stem - has a consonant onset and a vowel peak. The first type of syllable consisting only of one element is called a weak syllable, the one with consonant onset is called a strong syllable. Now what happens to weak syllables is that they often are simply dropped. This does not occur at the beginning, but in the middle of the sentence weak syllables (except aand nasals n-, m-) are being dropped as a rule.
Student 2:
Well, I guess that means that the tone of the weak syllable is also dropped?
Teacher:
In Africa, tones are tenacious of life. They tend to survive when the segments or even the words to which they were attached disappear. One could call this survival of the tones Christaller's Law, because it is Christaller who first discovered it more than a hundred years ago (Christaller 1893, Bearth 1994).
Student 2:
If it survives, what does it do? How does it survive if it is not pronounced? And how do we know that it survives?
Teacher:
There are essentially two ways in which a floating tone can survive. Either it finds a new 'host', that is it attaches itself to a neighbouring word or syllable which has not been dropped.
Student 2:
But these syllables have already their own tones, haven't they?
Teacher:
Very often yes. Then the newcomer must negotiate its place with the tones already installed.
Student 3:
And what is the other trick by which a floating tone survives?
Teacher:
It materialises as a Downstep on the following High.
Student 3:
How that?
Student 2:
I think I get the idea. The Low of the weak syllable of the word 'hat' disappears. But before it disappears it causes the following High tone to be lowered. So the Downstep is there as a memorial for the Low tone when it is no longer pronounced itself.
Student 3:
Or one could just as well say that it survives through the downstepping of the following High.
Student 1:
So this is exactly the Downdrift principle. The difference is simply that it is being triggered by a phantom tone.
Student 2:
Floating tone we said.
Teacher:
That is wonderful. Do you agree now that our example does not contradict the Downdrift Principle? Quite the contrary, it confirms it. As you see, the Downstep occurs exactly where an underlying Low can be reconstructed, that is where Downdrift would have caused a High to be lowered if the Low would still have been pronounced. 

 
ñà
L
H
L
!H
H
H
L??
Down-step 
"Give me my hat!"

-> spectrogram with analysis

Student 1:
I did not imagine that linguistics can be so dramatic. We have learnt all about the dramatic life of tones.
Teacher:
Not yet all. There is more to come.
Ghanaian friend:
I will tell my family about the life of tones.

-> Spectrograms (NP 2), -> Downstep and tonal syntax (NP 3)

Unit-6 Next