G. Oscar Russell

Mechanism of Speech

Past experiments have been confined mostly to an examination of the recorded speech curves or of speech phenomena after the sound emits from the vocal passages. And the physiological cause was imagined. Various conflicting vowel theories have resulted.A new method of attack has been carried out in the present investigation. Wheatstone harmonic or Willis cavity tone (erroneously called Helmholtz) theories are checked upon by direct physiological processes. Not only the forms but the exact dimensions of the speech cavities have been ascertained by means of X‐rays, palatograms and larnygoperiskopik measurements, made simultaneously and instantaneously so that exact measurements of the vocal tract resonators are had in three planes making it possible to reconstruct these cavities with a high degree of accuracy. By attaching an artificial glottis to the model it should be theoretically possible to reproduce the vowel if the total air volume functioning as a resonator were responsible for vowel quality. I am so...

Movies of Vocal Cord Speech and Voice Quality Vocal Cord Function

These are the first moving picture experiments to be obtained, bearing on the fundamental problems involved in a consideration of speech and voice quality vocal cord function.Most of our theories of speech have heretofore assumed a harmonic complexity of tones produced by the actuator or vocal cords. The moving picture photographs of the heretofore inaccessible interior larynx interior and lower throat now show for the first time that this is an impossible postulate.Speech and voice quality differences have heretofore been ascribed to a function of the total air volume cavity capacities above the vocal cords—in other words to a cavity tone modificatory influence. These pictures likewise prove that these theories are untenable and that a large part of the voice quality and many vowel quality differences may be traced to a function of the interior larynx.

Pedagogical Hearing Tests of Large Groups

At Ohio State University over 4000 students are examined yearly at 90 per hour. The 4‐A test was found inadequate and indeed misleading. So the following were developed. First, a push‐button for each student, recording on a compound kymograf synchronized with another recorder which simultaneously indicated the pitch and intensity of the tone at the instant it was passed into the head‐set. Too slow and cumbersome, though very accurate. Second, the student indicated by marks on a paper how many times he heard a 2‐A audiometer tone interrupted from one to five times. This modification of the 4‐A technique proved entirely unsatisfactory. Third, six sound‐proof booths, with push‐button operating flash light on the operators desk. These are arranged in groups of three to five, superimposed one above the other. One operator follows six to ten lights quite efficiently. This is now modified by a “weeding out” test. Essentially the same, except that only two tones are tested—256 c.p.s. and 8192 c.p.s. Fifteen students tested every three minutes. All failing to respond at 15 sensation units above threshold are then sent in for complete test. Including latter, 1000 students are tested daily Freshman Week of the Autumn Quarter.

Synchronized X‐Ray and Oscillographic Speech Records

The synchronization of physiological and physical speech studies would seem to be the logical next step in analyses of their characteristics. Not only should a study be made of the physical characteristics as manifest in the oscillographic record, but the physiological position at any exact instant should also be known if a complete understanding is to be had. It is obvious that both records will have to be automatic. The x‐ray exposure time will have to be reduced to at least 1/120th of a second if a study of normal speech is to be made. Prolonged or sung vowels are not characteristic of normal speech. Consequently such x‐ray procedures have long since been discarded by careful investigators. Since motion picture x‐rays consisting of 1/120th of a second are not yet feasible, the instantaneous x‐ray exposure used in this study is chosen as the next best technique. Sound picture records likewise synchronized therewith serve as a check back on the final resultant.

The Consonant—An X‐Ray Analysis of Its Vocal Organ Modifications by Other Sounds

Since, as the Abbe Rousselot, Scripture, Panconcelli‐Calzia, Crandall and a large number of others have pointed out, consonant sounds are of very short duration, analysis of their physiological characteristics by means of the x‐ray demands an automatic and very rapid technique. This is especially true of the unvoiced occlusives. Particularly in English. Notably in intervocalic positions. Likewise in certain combinations where laws of assimilation operate. But some unvoiced fricatives are also of very short duration. Crandall gives the mean for “p,” “t,” “k” and the unvoiced “th” as 0.05 sec. But the actual consonant contact or characteristic entity is obviously less than half that length. That demands an x‐ray time of at least no more than 1/120th of a second. Also an absolutely automatic opening of the exposure taking place with contact or rather a fractional few thousandths of a second after this central phase position has been assumed. That we have been able to accomplish in this study. Consequently th...

Some Terms of Physics for Linguists

If we threw a cork out on the waves of a pond, it would rise and fall as they passed under it. From the time it rode one crest until it dipped down and came up on top of the next, would represent one double vibration, and the number of such dips it had to make in a second would show the speed at which the wave was travelling and could be represented in d.v./sec. So, if it made 100 such dips in a second, we would say the waves were travelling at the rate of 100 d.v./sec. It is a well known fact that sound waves travel thru the air in much the same manner the waves on a pond are propagated when you throw a pebble into it; that is they radiate in all directions from the point of stimulus. But unlike the pond waves which are confined to one flat surface, they are transmitted up and down as well as along a horizontal plane. If we could see one at any given instant it would appear much like a series of progressively larger balls one inside the other, except that they would be made up of alternate layers of condensed and rarified air instead of rubber walls, and each would be spaced at an equal distance from the other. Of course in either case it is what the physicist calls the energy of the wave which travels, and not the medium itself. The cork out in the middle of the pond will bob up and down in the same place as the gently undulating waves pass under it; and the same particle of water will remain in the same place regardless of the dips which pass along, all providing that no other factor such as churning, friction, capillary attraction, etc. be involved. So in Speech. Consequently we must not lose sight of the fact that these vibrations we talk of are but energy vibrations. They are of the type of shocks we observe if we see an engine give a long string of freight cars a bump when the one farthest away is cut loose and not coupled on but merely resting against the string; the string stays where it is, but the last car goes shooting down the track in consequence of the