TY - JOUR TI - Writing sound with a human ear: reconstructing Bell and Blake’s 1874 ear phonautograph AU -Tom Everett PY - 2019 VL - IS - Autumn 2019 KW - acoustic science KW - Alexander Graham Bell KW - Clarence J Blake KW - ear phonautograph KW - reconstruction KW - sound technology KW - telephone origins KW - visible speech AB - Museums collect, preserve and interpret physical objects because they contain historical evidence that is sometimes difficult – if not impossible – to access through text and image-based research alone. In cases where original objects have been lost to history, reconstructions offer one way of reclaiming aspects of a past object’s design, use context and material culture. This article describes one such reconstruction project, which I recently undertook with collaborators at the Canada Science and Technology Museum in Ottawa. The goal of the project was to reconstruct Alexander Graham Bell and Clarence J Blake’s ear phonautograph: an 1874 curiosity that used an excised human middle ear to visually inscribe sound waves. Originally conceived of as a tool for deaf education, it became better known as the instrument that provided Bell with the technical insights he needed to develop and patent the telephone in 1876. This article consists of three parts. The first provides a historical background of the ear phonautograph and rationale for its reconstruction; the second offers a detailed description of the research and fabrication process; the third presents preliminary research findings that emerged through our experience building and operating the completed reconstruction. N1 - The human middle ear is the section between the outer ear (auricle, external ear canal) and inner ear (vestibular system, cochlea). It is comprised of the tympanic membrane, or ear drum, and ossicle bones (malleus, incus, and stapes). N1 - For more on Édouard-Léon Scott de Martinville’s early phonautographic experiments, see Pantalony (2009, pp 41–47), and Feaster (2010). N1 - Preservation in this case refers to the preservation of visual inscriptions of sound, not the sounds themselves. This said, historians David Giovannoni, Patrick Feaster, and collaborators on their ‘First Sounds’ project have recently been successful in using software to recover sounds captured in Scott de Martinville’s original phonautographic etchings. You can learn more about the project, and listen to the audio recordings, here: http://www.firstsounds.org/research/scott.php N1 - An extensive English translation of Scott de Martinville’s known phonautographic manuscripts (c. 1850s–60s), which outlines these and other technical details of his instruments and experiments, can be found in Feaster (2010). N1 - For an overview of specific instruments that preceded and informed the ear phonautograph’s development – among them Scott de Martinville’s standard phonautograph (made by Rudolph Koenig), Koenig’s manometric flame apparatus, Charles Anson Morey’s modified phonautograph, and Adam Politzer’s experiments using human and bird ear specimens to graphically illustrate sound (which Blake describes as constituting ‘the first record of the phonautographic use of the membrana tympani’) – see Bell, 1877, pp 404–405; Blake, 1878, pp 458–60; Bell, 1879, pp 172–173; The American Bell Telephone Company, 1882, pp 114–123; Snyder, 1974, p 11. N1 - Upon reviewing the various tympanic membranes employed in earlier phonautographic instruments, among them ‘thin bladders, gold-beater’s skin, thin rubber cloth, or even parchment paper and collodion film’, Blake concluded that ‘differences consequent on changes in pitch and on varying degrees of tension show more plainly in the almost microscopic tracings obtained from the membrana tympani than in the tracings made by the coarser mechanical device of the mechanician’ (Blake, 1876, pp 109–110). See also The American Bell Telephone Company, 1882, p 337. N1 - ‘[I]t was thought that my father’s system of pictorial symbols, popularly known as visible speech, might prove a means whereby we could teach the deaf and dumb to use their vocal organs and to speak’, wrote Bell in 1877. ‘The great success of these experiments urged upon me the advisability of devising methods of exhibiting the vibrations of sound optically, for use in teaching the deaf and dumb’ (Bell, 1877, p 404). N1 - For more on the relationship between the ear phonautograph and visible speech, see Snyder, 1974, pp 11–12; Sterne, 2001, pp 266–267; Mills, 2010, pp 38–47. N1 - Although the ear phonautograph proved functional, Bell ultimately found the etchings it produced to be inadequate for the purposes of vowel study. As he wrote in 1879: ‘I found it impossible to recognize the various vowel sounds by their tracings’ (Bell, 1879, p 173). See also Mills, 2010, p 47. N1 - See Bell, 1877, p 406; The American Bell Telephone Company, 1882, pp 114–123, 335–350; Bruce, 1973, pp 120–122; Snyder, 1974, pp 14, 30; Sterne, 2001, p 266; Mills, 2010, p 38. N1 - My use of the word ‘object’ in the singular is a slight misnomer, as Bell and Blake in fact built two ear phonautographs: one for Bell to use in Canada, and another for Blake to use in the USA (Snyder, 1974, p 12; Sterne, 2001, p 270). Both instruments were devised in collaboration between Bell and Blake, are believed to have been nearly identical in form and operation, and there is no evidence to suggest that either has survived over time. To avoid unnecessary confusion that might arise through the simultaneous discussion of two ‘original’ Bell and Blake phonautographs, I have opted to refer to the instrument(s) in the singular. N1 - For example, while Bell is sometimes construed as a champion of deaf education in relation to his work on visible speech – and by extension, his work on the ear phonautograph – the reality is more complex. As deaf scholars have long shown, the purpose of visible speech was not to provide support for the broader deaf community on their own terms, but instead to force deaf students to assimilate into hearing culture by learning how to read lips and speak clearly: a programme known as oralism (see Sterne, 2001; Greenwald, 2007; Mills, 2010). At the time of the ear phonautograph’s development, Bell was already an active critic of emergent sign-based language systems and segregated deaf schooling systems, and later became a vocal proponent of eugenics as a means of preventing ‘the formation of a deaf variety of the human race’ (Bell, 1884, p 4; see also Greenwald, 2007). By drawing attention to the controversial origins of the telephone vis-à-vis the ear phonautograph, I hoped our display might contribute to a more nuanced and critical discussion of this historic achievement. N1 - See also Heering and Müller, 2002; Heering and Sichau, 2005; Boon et al, 2017; and ongoing research by the Reconstruction, Re-enactment and Replication Network at Utrecht University (https://rrr-network.com/). N1 - The methodology we employed was broadly informed by Staubermann’s Reconstructions: Recreating Science and Technology of the Past (2011, p VIII), in which he describes four main elements common across many successful museum-based reconstruction projects: ‘the examination of existing artefacts and records, the construction of the replica, its use, and the interpretation of the experience gained.’ N1 - My use of the term ‘hacker’ is informed by authors such as Rosner and Bean (2009) and Kolko et al (2012), who describe how the term has been taken up by do-it-yourself, maker, and craft communities to refer to the creative modification of pre-existing technologies to create objects that achieve new, and often unforeseen, outcomes. As Kolko et al (2012, p 130) explains: ‘“Hacker” used in this context doesn’t necessarily mean a computer hacker who breaks into systems, nor does it refer to hobbyists; rather, the hacker identity combines a DIY ethic (i.e., hands on) with a passion for exploring the unarticulated potential of existing technologies.’ The concept of object hacking also has similarities to that of tinkering, as described – most often in relation to consumer electronics technologies – by authors such as Douglas (1987), Takahashi (2000), Pinch & Trocco (2002), and Waksman (2004). N1 - My use of the term ‘hacker’ is informed by authors such as Rosner and Bean (2009) and Kolko et al (2012), who describe how the term has been taken up by do-it-yourself, maker, and craft communities to refer to the creative modification of pre-existing technologies to create objects that achieve new, and often unforeseen, outcomes. As Kolko et al (2012, p 130) explains: ‘“Hacker” used in this context doesn’t necessarily mean a computer hacker who breaks into systems, nor does it refer to hobbyists; rather, the hacker identity combines a DIY ethic (i.e., hands on) with a passion for exploring the unarticulated potential of existing technologies.’ The concept of object hacking also has similarities to that of tinkering, as described – most often in relation to consumer electronics technologies – by authors such as Douglas (1987), Takahashi (2000), Pinch & Trocco (2002), and Waksman (2004). N1 - While other sketches of the original ear phonautograph do exist – such as a line drawing published by du Moncel (1879) – we didn’t want to infer too much information from these images as we couldn’t confirm their authenticity nor identify their original sources. Regardless, these images didn’t appear to offer anything dramatically different than those offered by Blake (1876) and Bell (1877), so we felt comfortable proceeding without them. N1 - Neither Bell nor Blake make any mention of the ideal heaviness of this weight, nor the preferred rate of speed at which the glass plate should travel, remarking only that the plate must be ‘passed rapidly’ (Bell, 1877, p 406), ‘smoothly and at a uniform speed’ (Blake, 1875, p 122). Our operational testing suggests that two to three seconds is likely the speed required to capture an optimally-shaped etching (see Research Findings section below), and further experimentation should help us determine the ideal weight for standardising this operation. N1 - Historians have established that Blake prepared at least two ear specimens for phonautographic use: one for use in Bell’s ear phonautograph, and one for his own (see note 11). It appears likely, however – upon a close reading of Blake’s published reports – that additional ears were also experimented with as he attempted to formalise a method for excising/preparing an ear specimen for this purpose. In his 1875 report, for example, Blake writes that ‘the [hair] saw should be inclined inward toward the stapes, and the descending process of the incus pressed gently outward by means of a spatula, in order to avoid any movement of the incus, a touch of the saw upon the long process of the incus being often sufficient to rupture the capsular ligament of the articulation of the incus and malleus, and so far derange the relations of the ossicula as to render the specimen unfit for experiment’ (p 122, my emphasis). N1 - 3D Print Exchange is an initiative led by the National Institute of Allergy and Infectious Disease, in collaboration with the Eunice Kennedy Shriver National Institute for Child Health and Human Development and the National Library of Medicine. Its aim is to provide ‘an open, comprehensive, and interactive website for searching, browsing, downloading, and sharing biomedical 3D print files, modeling tutorials, and educational material’ (https://3dprint.nih.gov/about). I was able to confirm, through an e-mail correspondence with the image creator/uploader, that the scan used in our project was created from a ‘thin-cut head CT of a patient undergoing endoscopic surgery for a skull base tumor’. I was also told that patient consent regarding the sharing of this image wasn’t requested, as the scan was modified and de-identified to remove any features that might reveal the patient’s identity – a practice consistent with the Health Insurance Portability and Accountability Act’s (HIPAA) legal guidelines for medical image sharing (see, for example, Harting et al, 2015, p 406–408). The ethics of medical image and data sharing are, at present, widely debated – particularly in the context of social media and open source platforms. For more on these issues, see Freymann et al, 2012; Harting et al, 2015; Crane and Gardner, 2016. N1 - As per Blake’s instructions, only two of the three human ossicle bones were fabricated and employed in our reconstruction: the malleus and incus. The innermost ossicle bone, the stapes, was considered extraneous to the phonautographic functioning of the ear (see Blake, 1876, p 111–112). N1 - In his own writing, Bell describes the stylus as being made of hay (Bell, 1877, p 405). N1 - Following Bell (1877, p 405), we opted to affix our stylus to the incus, though Blake left the question open as to which ossicle bone – the malleus or the incus – was in fact most ideally suited to this purpose (see Blake, 1876, p 112). N1 - Common nineteenth century microscopes are typically positioned with two ‘feet’ in front, and one in the back. In the ear phonautograph, the feet are positioned inversely: with one foot in front, and two in the back. The position of the base was likely changed in order to provide greater stability to the front of the instrument, as the addition of the moving glass tray apparatus – in our experience – made the instrument considerably more top- and front-heavy. N1 - Upon comparing the size of the circular mirror in our final reconstruction to the size of the mirror pictured in our ear phonautograph reference images (Figures 1 and 2), it seems most plausible that Bell and Blake used a larger microscope than the one we procured for our experiment. N1 - Returning to our reference images (Figures 1 and 2), we concluded that Bell had likely left the mirror intact simply because there was no reason to remove it: while it may not have served any functional purpose, we found no evidence to suggest that it would have impeded the instrument’s use in any way (indeed, we found no reason to go through the trouble of removing it in our own reconstruction). As for Blake, he most likely removed the mirror and repositioned the semi-circular bracket in order to make the instrument look more streamlined and accurate to its functionality. Since the purpose of his staged photograph was to convey – for readers of the scientific journal in which the photograph was published – the exact design, functionality and scientific application of the instrument, we reasoned that he likely took these steps to ensure that no extraneous components were visible that might distract from or otherwise confuse its operation. N1 - See note 15 above. N1 - Of course, this same resourcefulness and creativity led to other, less savoury, aspects of the ear phonautograph’s development, such as Bell and Blake’s rather cavalier approach to sourcing the human ear specimens required to build the instrument. The historical duality of the term ‘hacking’ – as having both positive and negative moral connotations – is perhaps instructive here, as a descriptor of activities that both ‘play upon the strengths’ and ‘take advantage of the weaknesses’ (Rosner and Bean, 2009, p 420) of existing technologies, systems and people. PB - The Science Museum Group SN - 2054-5770 LA - eng DO - 10.15180/191206 UR - https://journal.sciencemuseum.ac.uk/article/writing-sound/ T2 - Science Museum Group Journal