LB 107-110T

Ⅴ. THE PROBLEM OF THE ORGANIZATION PRICIPLE: RHYTHM
The production of speech and the understanding of language may be sustained for several hours without any interruptions longer than a few minutes. Within one minute of discourse as many as 10 to 15 thousand neuromuscular events occur. The facility for production or reproduction of this multitude of overlapping and closely timed activities cannot possibly have been acquired by a simple rote-learning procedure or by any other direct “stamping in” method (Miller, Galanter. and Pribram, 1960). There must be some organizing principle that underlies the perception of speech and language as well as the intricacies of timing and ordering during production. What is that general, organizing principle? Lashley, perceptive also in this respect, proposed a rhythmic phenomenon as an explanatory construct. He was, however, vague and had no empirical evidence to back up this working hypothesis. A similar solution suggests itself for reasons quiet unsuspected by Lashley.

語音和語言認知可以維持好幾個小時而不被打斷.在一分鐘的談話中,有一萬到一萬五千個神經系統的肌肉在運作.語音的流暢或是大部分重覆的部分和動作並沒有辦法運用簡單的機械式的方式呈現.當中必然存在著一些強調語言知覺和時間以及語音音序的器官組織原則.什麼是普遍的組織原則?Lashley認為,知覺作用也在其中.他提出旋律的現象和有助於解釋的構成理論.然而,他卻模糊不清,也並沒有提供根據觀察之後所得到的證據來支持其假說.而另一個相關的解決方式卻無法得到Lashley的認同.

The sequence of speech sounds that constitute a string of words is a sound pattern somewhat analogous to a mosaic; the latter is put together stone after stone, yet the picture as a whole must have come into being in the artist’s mind before he began to lay down the pieces. In the progress of his work, he may put down three contiguous stones, each of which may in the end contribute to the same or to unrelated pictorial units. When we talk about visual patterns we consider only spatial dimensions, disregarding the dimension of time. Under most circumstances, time does seem to be irrelevant. Yet, physiological processes do have a temporal dimension, and even in the process of seeing, which strikes us as taking place instantaneously, time plays a role. The identification of such simple figures as triangles and circles requires time and consequently, requires temporal integration in the central nervous system.

所謂聲音的模式,是由語音的連續性構成了一串文字.有點類似馬賽克磚.馬賽克磚是將石頭一個一個排列整齊,在整個作品完成之前,作者必須先在腦中構思.而在他製作的過程中,他可能會先將相鄰的三塊石頭排列在一起,每一塊石頭最終都是為了要能完成作品.當我們討論到視覺的模式,我們所考慮到的只有空間的範圍,而不討論時間.而在視覺的過程中,當我們看到一件突發事件.時間則扮演著重要的角色.例如簡單的圖像,三角形或是圓形的證明,需要時間來證明.因此,中樞神經系統則需要時間的統整.

In sound patterns, the entire configuration is in the realm of time and the problem then becomes: How does pattern or order in time differ from randomness or disorder in time. What is it that enables us to recognize and to reproduce a time pattern – any time pattern?

在聲音的模式中.聲音完整的配置和排列會產生問題:在時間上,隨機性和失序性在模式或者是秩序上有何不同?為何能讓我們分辨並且去重新製造一個時間的模式—任何時間的模式?

In music it is well-known that it is possible to recognize melodies by finger tapping, or even head nodding; and after listening to ten seconds or more of finger tapping, we can discriminate fairly well between random tapping and patterned tapping. Behaviorally, patterned tapping can be memorized and is recognizable and reproducible; random tapping is not. The essential nature of a time pattern is an underlying pulse or beat. In the extreme case of order in time, the simplest pattern is the unadulterated pulse such as the tick-tock of a metronome. We can complicate this simple pattern by temporal modulation such as skipping a beat regularly or introducing additional taps between beats where those extra taps must occur at fractional periods of the time unit. The underlying pulse is the carrier on which the rhythmic pattern can be “fastened.” It is its indispensable ingredient in much the same way as a figure may only be recognized against a ground. Notice that of all the information contained in a melody none is as indispensable for recognition as that concerning time. We may eliminate variations in pitch, loudness, or timbre and still recognize the melody, but if we destroy the internal temporal relationships without distorting the other variables, the melody becomes at once unrecognizable (cf. also Sachs.1953).

在音樂中,大家都知道可以用拍手或者是點頭來分別出旋律;在聽完十秒或更多一點的拍手之後,我們可以非常清楚地區分出是隨機性的拍手或者是具有模仿性的拍手.具有模仿性的拍手可以被記憶,區別,並且被重覆.而隨機性的拍手則沒有辦法.最主要的原因是因為時間的模式強調拍子或者是連續性的拍打.在時間秩序的極端例子當中,最簡單的模式就像是節拍器,以時鐘式的擺動正確地表示出曲子拍子的器具.我們會被這樣簡單的模式所混淆是因為對於時間的調整,像是在拍子和拍子中間跳過一拍或者是多加一拍,這樣會造成在一個完整的時間單位中製造一些破碎的片段.特別被強調的拍子,就像是一個載體,可以使旋律的模式固定.要注意的是,任何一個有關於旋律的訊息都是應該注意的.我們或許會忽略拍子或聲音或音質的變化,仍然能辨別出旋律,但假使我們破壞了其中和時間的關聯,使其他的變因變形,則旋律就會無法辨別.

We have been talking here primarily of sound patterns. But our observations are actually applicable to all temporal patterns whether they are perceived through our ears, our eyes, our skin, or our sense of proprioception. In any medium, temporal pattern means a carrier pulse with modulations. Let us call the carrier pulse simply the rhythm.
我們在這裡已經討論了基本的聲音模式.但是我們的觀察要能確切地適用於所有的時間模式,不論是經由我們的耳朵,眼睛,或者是皮膚所感覺到的.在任何情形之下,時間模式是指可以調整拍子的載體,就讓我們簡稱為節奏.

If speech is a patterned temporal phenomenon, and if such phenomena are based on underlying rhythms, is articulation rhythmic?

如果講話是模仿時間現象,假使這種現象主要是以強調節奏為主,那麼發音是有節奏的嗎?

(1) The Rhythm Nature of Articulation

A rhythm may be marked by equidistant pulses or by simple oscillations. A special
case of the latter is the periodic alternation between two states, say sleep and wakefulness, or facilitation and inhibition. The rhythm underlying speech seems to be based on rhythmic alternations between states although we cannot yet say what the origin or nature of these states might be. Because of our ignorance of the true physiological basis of these states, we must be content with thinking of them as purely theoretical constructs. Let us say they are states of initiation and execution of motor patterns (or cycles of activation and inhibition).

節奏可以用平行的節拍或者是簡單的擺動來表示.對於後者,存在著一個特別的情況,以週期性的交互作用存在在兩種不同的狀態之間,休眠和非休眠,或者是促進和壓抑.強調講話的節奏似乎是在這兩種不同的狀態下交互作用.雖然,我們現在還不能了解這些狀態的開端和本質為何.因為我們對於這些狀態在生理學基礎上的無知,我們必須要樂意地思考其開端以及原動模式的演奏技巧.(或者是活動和阻礙的週期)

An analogy might illustrate the point. Take once more the drummimg of our fingers upon a table top. We can make the taps follow one another in rapid succession and with practice we may learn to tap without introduction a longer pause or louder tap at the time we start with the small finger again. Nevertheless, the tapping is organized in terms of a single motor pattern of the hand; for every four taps we have to repeat that pattern. If we tap simultaneously with both hands we have two such pattern going on at the same time, and the individual taps from the right and the left hand will intermingle in their temporal sequence. Ordinarily we can hear the tapping rhythm (that is, the grouping by fours), but in some cases the rhythm may no longer be recognizable. However, even in these cases, there are statistical means by which the underlying rhythmicity could be demonstrated. Each motor pattern is somewhat similar to fundamental motor patterns that underlie speech, probably corresponding to syllables.

用類比的方式來說明這個觀念.我們在桌面上用手打鼓.我們可以一個接著一個連續地打拍子.在練習的時候,我們可以不用製造出太長或者是太大聲的拍子.尤其是當我們重新用小拇指練習的時候.雖然拍子是由一個一個手的原動模式所組成的,但我們仍能重覆一個四拍的節奏.如果我們同時用左右兩手,則左右手各個拍子會融合於他們的時間的連續性當中.正常情況,我們可以聽到拍子的節奏(四個一拍).但有些情況之下節奏可能無法辨識.然而,即使如此,那些節奏仍可用統計的方式來證實.每個原動模式多少都和那些已發音或者是相當於以音節為基礎的基礎原動模式有相類似之處.

Let us hypothesize that there is a basic periodicity of approximately six cycles per second.* Since we are not dealing with a mechanical device, we must expect some variations within and among individuals. It is, therefore, safer to hypothesize a rate of 6 +/- 1 per second. Thus, one-sixth of a second is taken to be a time unit in the programming of motor-speech patterns. With this assumption, a great variety of phenomena may be explained. The basic facts pertaining to them are well-established, but so far individual, unrelated explanations have been offered for each of them. A single hypothesis concerning an underlying rhythm brings them all together. We shall discuss each under a separate heading.

讓我們假設有一個大約以每秒為六個循環的基礎週期.因為我們並沒有使用機械器材,我們必須假設會有一些變量在其中.所以,比較安全的情況是假設平均每秒為6+/-1的頻率為單位.如此,1/6秒就視為原動發音模式的時間單位.在這個假設之下,多種不同的現象可以得到解釋.附屬於這些現象的基礎事實以完善建構.但對於個別,非相關性的解釋,仍相互挑戰中.有個和節奏相關的假說涉及其中.我們應該分節來加以進行討論.

(a) Delayed Feedback. Normally we hear our voice at practically the same time as it is produced. Speech may be seriously disrupted if a delay is artificially introduced between the time we actually speak and the time the corresponding sounds reach our ears. This phenomenon is often called the Lee effect. J. W. Black (1951) studied the relationship between the length of the delay and the degree of speech interference. He measured the latter in terms of the time it took subjects to read certain test material. In Fig.317 this variable is plotted as a function of the delay time. It appears that there is a critical delay that maximizes interference. The greatest interference occurs with a delay of about 180 msec(2/11 sec.). The curve found by Black is more or less what we would expect from the rhythm hypothesis.
* I am indebted to A. W. F. Huggins for valuable regarding the following paragraphs.

(a)延遲反應.通常我們聽到自己的聲音實際上就是聲音製造出來的時間點.如果延遲現象是刻意製造在我們實際說話和聲音進入到我們的耳朵之間,則語音有可能會變成片斷的.這種現象通常稱為李氏現象.J.W.Black(1951)研究延遲現象的長度和語音干擾的程度,他以有提到主詞的時間點測量後者,顯示出某些測試內容.在圖3.17以延遲時間的作用製成圖表,顯示出干擾到達最大限度的關鍵延遲現象.最大的干擾發生在180msec(2/11sec)中.而Black所發現的曲線我們多少是可以從有關節奏的假說中預知到的.