Steampunk phonetics

« previous post | next post »

From the Transactions of the Philological Society, 1873-74, "VIII. — On the Physical Constituents of Accent and Emphasis: By Alexander J. Ellis, Esq., President":

Phonautographic Sound-curves. Any disturbance in the air produces a series of alternate condensations and rarefactions, which, coming in contact with the drum of the ear, cause it to vibrate, in such a manner as to produce, after various internal modifications, the well-known sensation of sound. The most convenient way of analyzing this sensation is to analyze the motion of a single point in the drum of the ear. This is effected by an instrument called the phonautograph, consisting of a metal paraboloidal reflector (answering to the passage leading to the drum of the ear), truncated by a plane passing through its focus and perpendicular to its axis, over which opening is stretched a delicate membrane, ordinarily bladder (answering to the drum of the ear). At one point of this membrane is fixed a style (ordinarily a piece of quill), which rests against a cylinder, over which is rolled a piece of paper delicately coated with lampblack. A disturbance of the air inside the reflector causes the style to move backwards and forwards on the lampblacked surface, which it scrapes off. If the cylinder remain at rest, this produces a white straight line of moderate length. But if, as is usual, the cylinder be caused to revolve with a uniform motion, the style scratches out a white undulating line, which may be called a sound-curve, and which is the visible symbol of the invisible disturbance of the air.

I'll have more to say about this fascinating document, about Alexander J. Ellis (see also Jonathan P.J. Stock, "Alexander J. Ellis and his place in the history of ethnomusicology", Ethnomusicology 2007), and about Ellis's 1872 Philological Society Presidential Address ("On the Relation of Thought to Sound as the Pivot of Philological Research").

For now, I'll just add Ellis's list of "matters to be considered in a sound-curve", and his quaint account of sound-curve duration:

A sound-curve is generally made up of repeated undulations, and is considered to be unaltered so long as these undulations recur in precisely the same manner. There are four principal matters to be considered in a sound-curve, which will be here called length, pitch, force, and form, the nature of which I proceed to explain.

Means exist for accurately determining the length of the medial straight line, which would be described by a fixed, instead of a vibrating style, in one second of time, and would then pass from the beginning to the end of a sound-curve. The ratio of the length of such portion of this line as passes through an unaltering portion of the sound-curve, to the length corresponding to one second of time, gives the length of time during which the disturbance of air lasted unchanged, that is, the duration of the sound in seconds. This will hereafter be called its length.

I'll also note in passing that the term "Sound-curve" — what we would now call a "(sound) waveform" — did not die with Ellis, but was used as late as 1937 in Sir James Jean's Science and Music:

We have seen that all sound, whether pleasant or unpleasant, whether music or mere noise, is represented by a curve. We shall first examine the general properties of such a curve, trying to discover why some curves produce pleasure when they reach our ears and some pain. We must then consider the transmission of sound, discussing how best to retain the pleasurable qualities in our sound-curve, as it passes on from one carrier to another, and how far it is possible to prevent unpleasant qualities contaminating the curve. Finally we shall have to discuss the strange transformations that the sound-curve may undergo inside our heads.

As background, this is what a "phonautograph" device of Ellis's era (sometimes called "phonoautograph") looked like (source and discussion here):

And this is the sort of output it produced (source and discussion here):

A later (c. 1913) version of the same sort of device, from L'abbé Rousselot's lab at the Sorbonne, is illustrated in my post "Empirical Foundations of Linguistics", 7/26/2011. Rousselot's machine has a much longer piece of lampblacked recording material, a clockwork motor to drive the cylinder, and multiple inputs for things like nasal airflow:

(See also "God's own Englishman with a tube up his nose", 3/23/2007.)


  1. Guy said,

    August 23, 2015 @ 1:43 pm

    According to Wikipedia: "The phonograph was invented in 1877 by Thomas Edison. While other inventors had produced devices that could record sounds, Edison's phonograph was the first to be able to reproduce the recorded sound." (Internal citations omitted)

    All the accounts I had seen previously had made it sound like Edison was responsible for more than figuring out a way to reproduce it, one account said that Edison had not only created the device, but was the first to even think of the possibility of recording sound. I suppose I shouldn't be surprised to learn that these accounts are heavily embellished and romanticized.

    [(myl) For a fascinating and surprising account of Edison's path to the invention, see Patrick Feaster, "Speech acoustics and the keyboard telephone: Rethinking Edison's discovery of the phonograph principle", ARSC Journal 2007:

    …a closer look at the documentary record sugests that Edison was neither well versed in Helmholtz's acoustic theories nor aware of the phonautograph …
    Here I will pursue a different approach to the question of how Edison came to discover the phonograph principle. […] I find that […] the main impetus leading Edison to conceive of the phonograph principle probably came from his effort to develop a prospective invention that past historians of the phonograph have entirely overlooked: the keyboard telephone. […]
    Edison brought a formidable understanding of electricity and chemistry with him to the field of telephony, but there is evidence that his knowledge of acoustics was initially quite weak. […]


  2. david said,

    August 23, 2015 @ 2:55 pm

    The smoked drum kymograph, the recording part of the phonoautograph, was invented in the 1840s and was still in use in university student labs in the 1960s. Paper chart recorders with ink pens are still in use but are mostly replaced with digital devices now.

  3. John Lawler said,

    August 23, 2015 @ 3:04 pm

    > If the cylinder remain at rest, this produces a white straight line of moderate length. But if, as is usual, the cylinder be caused to revolve with a uniform motion, …

    Wow! The present subjunctive was alive and kicking in scientific discourse in 1874.

  4. a George said,

    August 23, 2015 @ 5:10 pm

    Ellis was slightly less than precise in the above, because the way he describes it, the uniform rotation of the cylinder would only have permitted one revolution of scratching. The trick was that the cylinder shaft was a threaded rod that passed through a nut, so that it not only rotated but also shifted axially, creating a helix for a baseline.

    A recent account of early graphical methods may be found in George Brock-Nannestad, "The mechanization of performance studies", Early Music, Vol. xlii, No. 4 (2014). The author makes reference to his English translation (the first in 150 years!) of the patents of Léon Scott who invented the phonautograph.

  5. Chas Belov said,

    August 23, 2015 @ 5:43 pm

    As opposed to now or as opposed to earlier?

    I find the wording odd, but does changing it to the following actually get rid of the passage's subjunctiveness? (Subjunctivity?)

    If the cylinder remains at rest, this produces a white straight line of moderate length. But if, as is usual, the cylinder is caused to revolve with a uniform motion, …

    [(myl) For why your example, whatever its subjunctivity, doesn't actually use the morphological subjunctive mood that John Lawler was celebrating, see "'Losing' the 'subjunctive'", 7/11/2004; "Linguistic therapy", 4/21/2011; and especially "Blithering idiocy on the subjunctive", 6/27/2012;

  6. Chas Belov said,

    August 23, 2015 @ 5:44 pm

    Oops! Interspersed post. My last comment was @John Lawler.

  7. Sidney Wood said,

    August 24, 2015 @ 5:00 am

    The inventor of the phonetograph was Édouard-Léon Scott de Martinville in 1857. The result was a waveform curve that couldn't be played back. His device is also illustrated in Franz Josef Pisko, Die neueren Apparate der Akustik, Vienna 1865. has details of Édouard-Léon Scott de Martinville and his work. They have also succeeded in playing back his waveforms.

  8. Sidney Wood said,

    August 24, 2015 @ 5:03 am

    That should have been phonautograph.

  9. a George said,

    August 24, 2015 @ 10:28 am

    Pisko's book is highly recommended; he analysed the phonautograph in great detail and was probably at the time one of the few to understand it. The Parisian scientific instrument maker Rudolph Koenig took out a license to Scott's French patents, but he did not develop the apparatus the way Scott wished. Scott did not renew his patents, which cannot have been very satisfactory to Koenig.

    It is ahistoric to call the inventor Éduard-Léon Scott de Martinville, because he only adopted the last part "de Martinville" late in his life. Much like saying that the scientific papers by John William Strutt before 1873 were written by Lord Rayleigh.

RSS feed for comments on this post