For decades, my day job gave regular need for bridging the gaps between the academically derived terminologies used for the labeling and classification of musical instruments in museum collections, the craft-oriented vocabularies of musical instrument makers, and the freer glossary used by musicians. I was deeply embroiled in what remains lively controversy about classification systems and am finding it increasingly difficult to steer clear of that topic on this blog.
At the moment, though, it seems to be something of a “Patent of the Month Club.” The nomenclature applied to the description of musical instruments in the reported documents varies widely and wildly, and is often severely at odds with that accepted in explicitly music-oriented contexts. Dealing with this is keeping the terminologist in me happily occupied. The present installment also provides a springboard into the discussion of tuning and tuning systems, which is another topic that I’ve been saying less about than I ultimately intend to.
Before getting to it, some of this blog’s followers may wish to note that I’ve recently edited last month’s post about overlapping patent claims fairly extensively in light of one that I had previously overlooked (for reasons not entirely unrelated to the introductory theme of the following discussion). The shield bars that define the Phonoharp were not an American invention datable to 1891. They appeared in an earlier German patent issued in 1887.
The heterogeneous vocabulary used in patents reflects varying approaches to the description of an innovation in terms that maximize the scope of the sought protection and the likelihood of its being granted, while minimizing the risk of a later successful challenge. Even if the employed terms of art originate with the inventor, misunderstanding and mistranslation is not uncommon with patents based on ones initially written in a foreign language.
This is well illustrated by German Imperial Patent No. 75089 issued to Julius Müller on 19 August 1893 for a “Device for positioning the moveable damper bars on chord zithers” (Vorrichtung zum Einstellen der verschiebbaren Dämpferleisten an Accordzithern).
He patented this design in several other countries. The one issued in the UK shortly thereafter was titled “Improvements in or relating to zithers” and names Müller as a “Zither Accordeon Manufacturer.” A second UK patent from the following year, for a closely related design and bearing the same title, refers to the instrument explicitly as a “zither accordion.” The US patent for the initial German design (no. 523902 issued on 31 July 1894) goes one step further and is headed “Zither-Accordion.” Although if seen in isolation this might appear to be a simple mistranslation of the German Accord-Zither, it clearly was not.
It may have been an artful hedge against potential patent conflicts that will be discussed in detail separately. However, the label zither-accordion can be derived on a purely lexical basis. Just as the first attested use of the term ‘autoharp’ is found in a well-known US patent tersely headed “Harp” and issued to Charles Zimmermann in 1882, the term ‘accordion’ can be traced back to an Austrian patent for an instrument bearing that name, issued to Cyrill Demian and Sons in 1829.
This incorporated the free reeds and bellows that were to remain characteristic of all varieties of accordion, but only produced chords. The instrument illustrated in the patent has a single keyboard with five buttons. Each produces one chord when the bellows are drawn open and another when they are pushed closed. The text notes that, if the added weight is deemed tolerable, a second keyboard can be placed on the other side of the bellows either for additional chords or single notes.
The Accordion was named for its production of chords and intended to provide immediately accessible means for accompanying music — just as the autoharp was. In that light, and with both instruments having button-activated mechanisms, referring to an autoharp as a zither accordion is as reasonable from the conceptual and nomenclatural perspectives as is the initial German designation Accordzither.
Numerous patents weighed into the accordion’s further development, with varying degrees of true innovation. Zimmermann participated with a small cluster of patents issued around 1870. He illustrated what would now be called a one-and-a-half row diatonic button accordion with four basses, in US Patent no. 106018 from that year.
The outer row of buttons on the melody side is tuned to a C major scale. The inner row adds a Bb and F# to each octave, supporting F and G major. (Zimmermann claimed the added row as his own invention despite citing a preexisting instrument that included it, in another of his patents.) Diatonic button accordions now commonly have one or two full diatonic rows on the melody side, sometimes adding a half or full third row. The two-row models typically have eight buttons on the bass side.
The selection of keys and the interval between the melody rows varies depending on the intended musical context. One widespread configuration separates them by a fifth. Another sets them a semitone apart, making a full chromatic scale available for melodic use, albeit with the tonality of the instrument still determined by the diatonic row that matches the bass side. Again as with autoharps, performers commonly have several instruments at hand.
The double stringing that now characterizes a diatonic autoharp or mandolin-strung chord zither, has a direct parallel on an accordion. A single button can open the air flow to one or more sets of reeds. Smaller diatonic instruments commonly have two such ‘registers’ or ‘stops,’ nominally tuned in unison. However, a slight pitch difference in the actual tuning of each pair of reeds causes a ‘tremolo’ effect that is carefully quantified and exploited. In addition to being a potentially desirable musical effect, it masks the significant impurity of the major thirds in the equal temperament to which such instruments are tuned.
A tuning that produces no tremolo is termed ‘dry’ and one where it is markedly present is ‘wet.’ An intermediate effect is attained by ‘swing tuning’ and there are several other named degrees. The pulsating increases in amplitude that constitute the tremolo are measured in ‘beats per second.’ That number is determined by the difference between the frequencies of the two reeds that are sounding together.
If one reed is tuned to 440 Hz and the other to 441 Hz, the sound will beat once per second. The difference between 440 Hz and 442 Hz is heard as two beats per second, 440 and 443 as three, and so on until the beating becomes too rapid to be perceived as a discrete component of the sound. Two reeds tuned in absolutely dry unison are beat-free and a wet tuning is about five beats per second; swing is one beat per second or a bit more. The beat rate doubles with each successively higher octave — 440/442 becomes 880/884 — but opinions vary about the need for tapering it toward the upper and lower ends of the instrument’s range.
Vibrating strings have the same property and everything said about tuning reeds on a diatonic accordion applies directly to, say, the double strings on a diatonic autoharp. Beats can also be heard between two strings slightly mistuned to consonant intervals such as major thirds, fifths, and octaves. Back in the days before electronic tuning devices had become ubiquitous, the determination of whether such an interval was in tune was commonly made by counting beats — zeroing them if a pure interval was desired and tempering an interval by setting its beat count. This was normally done by comparing the rates of adjacent intervals rather than in absolute terms. (For the technically inclined, the perception of this interaction involves several partial frequencies of vibration — aka overtones or harmonics — not just the fundamental. This is not a concern with the intrinsically harmonic vibration of reeds driven by a continuous flow of air but can be prominent with the inharmonic vibration of plucked strings.)
As long as one of the strings in each unison pair on a diatonic autoharp is accurately tuned, small imperfections in the tuning of the other strings can enhance the instrumen’s sonority. However, their random variation can as easily heighten the sense of out-of-tuneness, especially if the mistunings are on both the flat and the sharp side. The dry-swing-wet method provides means for controlling this to suit individual preference. Anyone curious about its practical utility can easily give it a whirl using an electronic tuner with an adjustable reference pitch.
Tune the instrument as usual and then reset the reference pitch to two hertz above what it was; if A=440, then to A=442. Retune the upper string in each pair. This will result in a pronounced swing effect (peculiarly labeled ‘demi-swing’). If it seems too extreme, reset the reference pitch to just one Hz above the initial one, and retune the upper strings beginning at the treble end. Test along the way and stop when the balance seems right.
A wetter effect can be attained by setting the reference pitch still higher. If the tuner permits it to be set in smaller than full Hz increments, an 0.5 Hz adjustment can make a noticeable difference. Retune the upper string in each pair beginning from the bass side, again testing along the way. If two equally reliable tuning devices are available, one can be set to the higher reference pitch and each string tuned in order with the devices side by side.
The banner image appearing above this post was taken from an advertisement in the 11 June 1901 issue of the Zeitschrift für Instrumentenbau. The full page where it appears is here.
The varying audible effects of accordions tuned from dry to wet are demonstrated in this video.