Typewriter Examination: Mechanical, Electric and Electronic

Mechanical typebar typewriters operate on a shared principle: each key depresses a typebar, a curved lever carrying a metal type slug at its tip. When the key is pressed, the typebar pivots upward (on most designs) or forward (on some European vertical-stroke designs) to strike the platen through the inked fabric ribbon. The force of the strike, the quality of the ribbon ink, the alignment of the typebar guide, and the condition of the type slug all influence the character impression on paper.

Class characteristics are shared by all machines of a given model or typeface design. The typeface itself is primary: the specific letterforms, serif style, stroke width ratios, and spacing of a given type slug set distinguish, for example, Pica 10-pitch type on an Underwood from Elite 12-pitch type on a Royal. Pitch (characters per horizontal inch) is a gross class characteristic: 10-pitch (Pica), 12-pitch (Elite), and proportional spacing (found mainly on IBM Executive and Varityper machines) divide the population. Typeface design was often proprietary: Remington's "Remington Standard" roman differs in specific proportions from Royal's "Royal" roman and from Olivetti's "Olivetti" roman, even though all three are superficially similar serif faces. Type catalogues compiled by manufacturers and by document examination reference collections (including the FBI Laboratory Questioned Documents Unit typewriter reference collection and the BKA collection in Wiesbaden, Germany) allow class attribution from impression characteristics alone.

Individual characteristics develop over the machine's service life. Out-of-alignment characters are the most common: a typebar guide slot that has widened with use, or a bent typebar, allows the type slug to strike slightly off its nominal position. The deviation is consistent for that key on that machine (a misaligned typebar consistently prints low-left) and distinguishes the specific machine from all others of the same model. Defective type is the second major category: chipped serifs, filled counters (the enclosed space of a letter "e" or "o" plugged with ink and debris), cracked or broken type faces, and worn type impression edges all leave characteristic damage impressions reproducible across documents typed on the same machine. Ribbon defects add a third layer: fabric ribbon striation patterns, ribbon reversal artefacts, and age-related ink depletion signatures can correlate typewritten documents to a specific ribbon spool or ribbon cartridge.

IBM introduced electric typebar machines in the early 1930s, adding motor-driven uniform keystroke force to the typebar mechanism. The forensic implication is that electric typebar impressions tend to be more consistent in ink density across a document than manually typed text (where finger pressure variation produces visible lightness and heaviness differences), but the underlying class and individual characteristics are otherwise similar to mechanical typebar machines.

The IBM Selectric, introduced in 1961 and produced through various models until 1986, replaced the bank of typebars with a single rotating spherical element approximately 4 cm in diameter, carrying 88 characters on its surface. Selecting a character involved tilting and rotating the ball to the correct position before it struck forward against the ribbon and platen. The element was interchangeable: a Selectric owner could switch among Pica, Elite, Prestige Elite, Letter Gothic, Courier, and other elements within seconds, giving one machine multiple typefaces.

From an examination standpoint, this introduces an important class-characteristic complexity. A document typed on a Selectric in Courier 12-pitch is not machine-identifiable from typeface alone because thousands of Selectrics used Courier elements. The examiner must focus on ribbon wear patterns (the Selectric's correctable ribbon cartridge leaves characteristic tape-lift patterns on correction stripes), platen condition (worn platens compress unevenly, producing consistent impression depth variation), and any physical damage to the specific type element (scratches, deposits, or mechanical deformation of individual character positions on the ball). The FBI's questioned document unit documented multiple cold-case typewriter identifications using Selectric element damage patterns during the 1970s through 1990s.

In the UK, the Forensic Science Service developed a type-element reference collection before its 2012 closure. Private practitioners in the US (certified by the American Board of Forensic Document Examiners, ABFDE) and in Germany (certified by the German Society for Forensic Document Examination, DGFDE) continue to use manufacturer catalogues and physical reference collections to attribute Selectric-typed documents.

Daisy-wheel printers and typewriters, commercially dominant from approximately 1978 to the mid-1990s, carry type on flexible plastic spokes radiating from a central hub, resembling a daisy or wagon wheel. The mechanism rotates the wheel to position the correct spoke under a print hammer, which strikes the spoke through an ink ribbon onto paper. Brother, Smith Corona, Olivetti, Diablo (a division of Xerox), and Qume were major manufacturers. The plastic spokes wear, crack, and deform in ways that differ from metal typebar defects.

Class characteristics include the daisy-wheel element designation (a specific font on a specific hub design, often labelled on the wheel itself) and the pitch encoding (usually 10 or 12 cpi, with proportional spacing available on higher-end elements). Because daisy-wheel elements were interchangeable (standardised hub sizes allowed cross-brand use of elements in some product lines), typeface attribution alone does not identify a specific machine model without corroborating evidence.

Individual characteristics from daisy-wheel machines include: hairline cracks in plastic spokes that produce ink gaps in specific strokes; chipped spoke tips that cause character truncation; spoke bending that introduces consistent vertical misalignment for specific characters; and hammer-force irregularities producing consistent light or dark character sets. The fabric or film ribbons used in daisy-wheel machines (carbon film ribbon was common for high-quality single-strike output; multistrike fabric ribbon was used for economy) add their own evidential layer. Carbon film ribbon lifts can sometimes be recovered from ribbon cassettes, revealing the actual text typed on the machine. The Forensic Science Regulator's Case Assessment and Interpretation guidance (UK) notes that ribbon recovery from carbon-film cartridges has been used in fraud and intellectual property theft investigations.

The primary examination tool is a stereo zoom microscope at magnifications from 5x to 40x, allowing the examiner to observe character impression depth, ink spread, type-slug edge definition, ribbon impression texture, and alignment. A comparison microscope allowing side-by-side juxtaposition of a questioned document and a known standard from a suspect machine is the gold standard for machine-level identification, mirroring the use of the comparison microscope in ballistics examination. The specialised imaging tools of forensic physics, including UV and IR imaging modes, complement the microscope-based examination by differentiating ribbon inks that appear identical under white light.

Measurement of pitch (characters per inch) is performed with a calibrated transparent overlay or a stage micrometer, measuring the centre-to-centre distance across a known run of characters. Even within a single pitch class (all nominally "10 cpi"), manufacturing variation and mechanical wear create measurable pitch-spacing differences at the second decimal place. A Remington Noiseless 9 and an Underwood Standard both nominally produce 10-cpi output, but their actual spacing differs enough to contribute to class differentiation.

Vertical alignment is assessed by drawing a baseline reference through the foot of a series of characters and measuring deviation from the reference in tenths of a millimetre. A character consistently 0.3 mm high is a different defect signature from one that is 0.3 mm low; direction of displacement is part of the individual characteristic. Rotational misalignment (a character tilted clockwise or counterclockwise on the typebar) produces a consistent angular deviation measurable under the microscope.

Ink analysis complements mechanical examination. Infrared reflectance spectroscopy (using instruments such as the VSC6000, Foster + Freeman) can distinguish carbon-based fabric ribbon inks from nylon-spool inks, and can reveal alterations where different ink types were used at different points in a document's history. The spectral examination modes for detecting such alterations are described under detection methods: oblique, transmitted, UV, IR, and VSC. In the US, the FBI Laboratory and the Secret Service Forensic Laboratory maintain typewriter ribbon ink databases. In India, the CFSL (Central Forensic Science Laboratory) uses thin-layer chromatography for ink comparisons on typewritten documents, particularly in land-record fraud cases.

In the United States, the FBI's examination of the "Killian documents" in 2004 (documents purporting to be 1972 Texas Air National Guard memos about President George W. Bush's military service) relied partly on typographic analysis. Examiners noted that the documents showed proportional spacing and superscript characters consistent with Microsoft Word's default formatting, not with 1970s IBM Selectric output. While the Selectric offered some proportional-spacing elements, the specific character forms and formatting details matched contemporary word-processing output rather than period-correct typewriter production. This case, ultimately resolved on digital-forensic grounds, illustrates how typewriter examination must account for the full mechanical and temporal context of a claimed document.

In the United Kingdom, typewriter examination featured in several major fraud trials during the 1980s involving backdated tenancy agreements and company board minutes. The Forensic Science Service's document examination unit used typebar alignment and ribbon defect evidence to link series of fraudulent documents to single typewriters, even when the machines themselves were not recovered. The approach followed the framework codified in the ENFSI (European Network of Forensic Science Institutes) guideline on questioned document examination.

In India, disputed property and revenue documents typed on standard government office typewriters (Remington Rand and Godrej typewriters were the dominant makes in state government offices through the 1990s) have been a major source of typewriter examination casework at the CFSLs. The examination framework follows the same class-then-individual logic but with a reference population drawn from Indian government office typewriter inventories rather than Western commercial markets.

Forensic document examination follows a range of conclusion levels endorsed by the Scientific Working Group for Forensic Document Examination (SWGDOC) in the US, the ENFSI Quality and Competence Committee in Europe, and equivalent bodies in Australia (ANZPAA) and India (NABL). The levels move from positive identification (the document was typed on this specific machine) through probable identification (highly likely, minor unexplained variations present), inconclusive (neither identified nor eliminated), probable elimination, and positive elimination.

Positive elimination requires that the questioned document shows characteristics inconsistent with the known machine: for instance, a class characteristic (pitch or typeface design) that does not match, or an individual characteristic present in the questioned document that is absent in the known standard and cannot be explained by machine change or repair. A 10-cpi typeface on a questioned document eliminates a 12-cpi known machine without ambiguity.

Inconclusive opinions arise when the questioned document has insufficient character samples for reliable alignment measurement (a one-line handwritten note with typewritten header provides limited data), when the known standard is not contemporaneous (a machine repaired between the questioned document date and the examination date may have different alignment characteristics), or when image quality is insufficient for microscopic measurement (a photocopy of a photocopy of an original document).

Reporting in this discipline follows the SWGDOC Standard for Examination of Typewritten Items (last revised 2015) and the ENFSI Best Practice Manual for Document Examination. Both require that reports describe the examination method, the reference material used, the specific characteristics observed and measured, and the basis for the conclusion in terms the trier of fact can evaluate. UK courts applying R v. T (2010) admissibility considerations note that numerical likelihood ratios should not be used unless a firm statistical base exists, meaning typewriter identification opinions are expressed in qualified verbal terms rather than numerical form, given the absence of a population frequency database comparable to those supporting DNA or fingerprint examination.