Reconciling Conflicting Opus Dates, Builder Names, and Institutional Records

Records disagree most often because they were never answering the same question. The Organ Historical Society describes its Pipe Organ Database as a dynamic database whose content is always changing and will always be in a state of flux, so a database record read today is a dated snapshot of what a contributor entered, and the instrument and builder counts it cites for its legacy import are not a current count. An archival finding aid answers a different question again, because its title dates describe the paper in the folder rather than the instrument the paper concerns, and a corporate filing or a parish page dates its own event, so a date carried by any of them can be a file date, a legal date, or a dedication date rather than the date of building.

A date in a record is evidence of the record’s own event first, and Estey Opus 89 shows how much that matters. The Organ Historical Society’s archival-object title for the Opus 89 file carries the span 1912 to 1913, while the published Estey opus list gives 1903 for the same opus; read as a file date span rather than a build date, the title most likely records when the paper was generated, which reframes the disagreement as a category difference rather than a factual conflict about when the instrument was built. That reading is an inference from the form of neighboring titles in the same collection, the repository does not state what its title dates mean, and the inference does not decide which date is correct. Both positions therefore stay on record: 1903 from the compiled opus list, which its 2020 revision still gives, and 1912 from the archival-object title, neither averaged nor silently preferred. They could describe different events such as a build and a later rebuild, but nothing on record says so.

An institutional record or an owner’s page can disagree with the builder’s own timeline about the same instrument, and the disagreement is carried rather than resolved. A Dordt University repository record for an illustrated history published in 2015 describes a three-manual tracker-action organ built by Casavant Frères in 1979 at Dordt College in Sioux Center, Iowa, with no opus number stated, while Casavant’s own timeline lists an Iowa installation of 1977 as opus 3360. It is not established that the two records describe the same instrument; if they do, the date differs by two years, and if they do not, two separate instruments are involved. A mirror of the Pipe Organ Database records the builder of the organ at St. Joseph’s Cathedral in San Jose, California as ‘J. H. & C. S. Odell & Co.’ with Opus 228 of 1886, while the cathedral’s own page attributes it to ‘J.H. & C.S. Odell Co., Yonkers, New York’ in 1886 with no opus number; neither is a primary factory document and neither resolves the name form or the address.

An opus number’s position in a builder’s sequence does not fix a build date, because the published pairs of number and year do not always run in order. Documented Kilgen pairs drawn from separate sources include Opus 3004 of 1919, Opus 4258 of 1929, Opus 5163 of 1933, Opus 5395 of 1935, Opus 5840 of 1927, Opus 5913 of 1938, and Opus 7722 of 1955, so Opus 5840 is dated eight years earlier than the lower-numbered Opus 5395. Each pair comes from a different source of varying reliability and some may simply be wrong, so the conclusion is that an opus number alone does not date a Kilgen without checking the opus list, not that Kilgen used a deliberately non-sequential system. Whether the inversions reflect Kilgen practice or a slip in a secondary source is left open, and no Kilgen company document stating a numbering rule is on record.

The anomaly is not peculiar to one firm: a regional practitioner’s compilation reports five Wicks opus and year pairs from 1973 to 1994 and one internally anomalous pairing, Opus 3447 with 1980, which is treated as unresolved rather than as evidence of a numbering policy. How any builder assigned numbers in general is owned by an opus-number page outside this group; this page enters an opus number only where two records disagree about it.

A single database record can disagree with itself, between the fields a reader searches and the document the record was built from. One Pipe Organ Database record for First Presbyterian Church in Morristown, Tennessee, read through an unaffiliated non-JavaScript mirror, carries a header reading ‘Schantz Organ Co.’, opus 1116, year 1973, while the stoplist attached to the same record is labeled ‘Op. 1306’ and 1974; that one record shows record-level conflict is one possible cause of an anomalous opus and year pair, and it does not quantify how often this occurs. Two entries in the same database can also disagree about one contract: for the same 1968 Möller rebuild at First Presbyterian Church in Washington, Pennsylvania, one entry is headed Opus R-786, 1968, while a 2024 note on the related entry describes the same work as R-768 in 1968. The difference is a digit transposition invisible to anyone who sees only one entry, which is why a prefixed number read from a secondary source is checked against the factory specification rather than repeated; the example establishes possibility, not prevalence.

One firm can appear under many name strings, and the strings are not interchangeable in a search. Across the sources encountered for the Westfield firm of William A. Johnson, the builder appears under at least six strings, ‘William A. Johnson’, ‘Wm. A. Johnson’, ‘William Johnson’, ‘Johnson Organ Company’, ‘Johnson & Son’, and ‘W. A. Johnson & Son’, so no single name-string search retrieves the firm’s records, instruments, or literature; that count is what was met in specific sources on the access date, not a census of every rendering in use.

The variation reaches primary records: United States patents filed from 1928 to 1933 name the assignee as Wicks Pipe Organ Co., a patent filed in 1939 names the assignee as Wicks Organ Co., and the OHS Pipe Organ Database catalogs instruments under the two forms as separate builder records. Even the founding name is contested, a tertiary encyclopedia article stating that the Wicks brothers incorporated as the Wicks Organ Company in 1906, a university public-history entry stating they incorporated in 1908 and giving 1906 for Opus One, and the 1928 to 1933 patents naming Wicks Pipe Organ Co., a corporation of Illinois. As of August 31, 2026, the present entity’s website footer reads ‘Wicks Organs LLC’ while a trade-journal directory listing and the firm’s own event listing use ‘Wicks Organ LLC’, so both forms are preserved and the state registry is where a legal name is settled before it is used as one.

A single letter can split a firm in two: the founding partner’s surname is spelled ‘Schwarz’ in the National Register registration form and in the Reuter company’s own history but ‘Schwartz’ in tertiary sources, so the original firm name appears as both Reuter-Schwarz Organ Company and Reuter-Schwartz Organ Company.

A government filing or a registration form is not decisive on its own, because filings, company histories, and reference works disagree about founding dates, incorporation dates, and named founders. As of August 31, 2026, sources agree that The Austin Organ Company was formed in 1898 by John T. Austin and his brother Basil and was operating in Hartford, Connecticut from about 1899, but the manufacturer’s history ties incorporation to an 1898 factory fire followed by relocation to Hartford in 1899, while a tertiary encyclopedia article states it was formally organized in 1898 in Boston, so only the agreed 1898 formation and circa-1899 Hartford operation are stated without qualification. Möller’s founding is given three ways: several sources say 1875 in Greencastle, Pennsylvania, one detailed reference history describes a Greencastle partnership named ‘Möller, Brenisholtz and Co.’ running from about 1877 to 1881, and the Organ Historical Society’s own finding aid for the company’s records says the company was ‘established in 1881’, so each date belongs to the event its source describes and no single founding year is published.

The Reuter Organ Company shows both a founder conflict and a filing-date conflict: the National Register registration form names four people as founding the firm in 1917, Earl Schwarz, Henry Jost, Adolph C. Reuter, and Albert G. Sabol Sr., while an encyclopedia article names three principals and treats Sabol, Reuter’s nephew, as a founding employee who headed engineering and design. On incorporation the same registration form gives September 15, 1919 while the encyclopedia article gives September 5, 1919; the Kansas filing itself could not be retrieved, and the registration form is a preservation document rather than the corporate filing, so September 1919 is as precise as the record allows.

Who bought a firm, when, and what they bought are questions that sources of similar standing answer differently, and a widely repeated date range can rest on no independent citation at all. As of August 31, 2026, sources place a change of ownership of Austin Organs, Inc. to Michael Fazio and Richard G. Taylor in the mid-2000s but disagree on the mechanism and the date, the manufacturer and a 2014 U.S. Senate press release saying the pair acquired the company in 2005 and The Diapason’s 2007 feature saying American Classic Organ Company, led by the same two men, purchased Austin’s assets in January 2006; neither current ownership nor corporate structure is inferred from either.

The custody of the Möller company’s papers after 1992 is the clearest case of three accounts, none of which squares with the others, carried as separately attributed and separately dated statements. First, M. P. Möller, Inc. went into involuntary bankruptcy and closed in April 1992, and on January 13, 1993 the trade name and assets were acquired by King of Instruments, Inc. of Chicago; Jonathan Ambrosino, writing as an eyewitness of the January 1993 liquidation auction, states that King of Instruments purchased the company name and archives. Second, an encyclopedic article says the name, archives, and customer list went to King of Instruments, which briefly traded as ‘Möller Organ Company’ before ceasing in November 1994. Third, the Organ Historical Society reported in 1995 that the Allen Organ Company of Macungie, Pennsylvania acquired the Möller name and papers, and the Society’s finding aid for the M.P. Möller Collection states that the collection was purchased at auction by Allen and donated to the OHS in early 1995. No source establishes whether the files sold in 1993 were a different body of paper from the contract ledgers now at the OHS or the same paper described differently, none establishes any relationship between the two companies, and the dates attach to the events each account describes rather than to a verified chain of title. Who owns a defunct firm’s name and who holds its records are separate questions with potentially different answers.

A circulated range can cite nothing that supports it: on a reading of September 2, 2026, an encyclopedia article’s four Holtkamp name periods, including ‘The Votteler-Hettche Organ Company (1903-1914)’, are all footnoted to the Holtkamp company’s own history timeline, which names no Votteler-Hettche phase and dates the Votteler Holtkamp Sparling name to 1903. Dated documents can bound a transfer without fixing it: the date on which Robert Hope-Jones’s business passed to Wurlitzer is given as April 1910 by a general reference account, May 1910 by a Puget Sound Pipeline article, 1910 by the Theatre Organ Fact Finder, and 1914 by a widely read enthusiast site. A patent application filed on July 28, 1910 already shows Hope-Jones at North Tonawanda assigning to Wurlitzer, which is inconsistent with a 1914 transfer, but the patent bounds the date without fixing the terms or form of the transaction, and no corporate instrument such as a bill of sale or court record is on record, so the four positions stand with the patent beside them.

A heading, a program title, or a handbook line is not an attribution, and the number nearest a builder’s name in a heading is not necessarily that builder’s number. A broadcast program profile is headed ‘1928 Skinner; 2008 Schantz organ, Opus 634, at the Rockefeller Memorial Chapel, University of Chicago, Illinois’, while the owning institution states that the builder is Ernest M. Skinner and the opus is 634, dated 1928, so Opus 634 is Skinner’s and not a Schantz number. In two broadcast-heading examples tested against independent venue records, the trailing opus number belonged to the first-named original builder rather than to the later builder named nearest the number; that is a two-example warning, not a rule about how such headings are written.

Professional publications carry the same hazard in smaller form: a professional association’s convention handbook of October 2025 heads an article ‘St. Jude the Apostle Parish | Wauwatosa — Schantz Organ Co., Opus 2291, 2009’ and describes the instrument in its itinerary as ‘2009 Schantz, III/40’, while the company’s own portfolio page lists the same instrument at 39 ranks.

Aggregate figures and end dates conflict between sources of similar standing as a matter of routine, so a total or a closure year quoted from one source is a position, not a fact. Published figures for Möller’s total output run from ‘more than 12,000 instruments’ in encyclopedic and local-history sources, through ‘almost 12,000 pipe organs’ in the OHS finding aid, to ‘approximately 11,850 organs’ in the 2023 OHS Press company history, while the highest opus number encountered is 11739, dated 1990; the range and the source wording are preserved and no precise total is chosen.

Closure dates split the same way: The Diapason states that Anna Hastings ‘closed its doors in 1935’ and that the Hook & Hastings company ‘was dissolved in April of 1937’, the Peabody Essex Museum finding aid says the firm ‘survived until 1936’, an Organ Historical Society page gives closure in 1936, and the Organ Clearing House states the Weston factory ‘was demolished after the company closed in 1929’. The variants may describe different events, ceasing to build, ceasing to trade, formal dissolution, and factory demolition, but none of the four sources says which; likewise Aeolian-Skinner’s end is 1971 in the Aeolian-Skinner Archives timeline and 1972 in an encyclopedia article and a third-party builder record, so its closing is written as 1971 or 1972 until a dissolution record settles it. Specifications conflict too: the American Theatre Organ Society’s record gives the Oklahoma History Center’s 1935 Kilgen Opus 5281 as four manuals and 15 ranks, one rank more than the museum’s own 14-rank description, and both counts are preserved as attributed with neither adopted.

One public database is not one authority, because its pages, registers, and counters can disagree with each other. As of September 1, 2026, the Pipe Organ Database’s about page calls it ‘the definitive compilation of information about pipe organs in North America’ while its related-websites page states that it ‘accepts information about instruments across the globe’. On the same date its landing-page counters and its sitemap enumeration disagreed in both directions, the sitemaps listing 294 fewer instrument URLs and 15 fewer builder URLs than the counters of 75,220 and 8,477 but 111 more location URLs.

As of September 2, 2026, the builder register is internally inconsistent on the Steer & Turner partnership, one entry giving ‘Steer & Turner [J.W. Steer], Westfield, Massachusetts, 1866-1879; Springfield Massachusetts 1879-1891’ and a separate entry giving ‘Steer & Turner, Westfield, MA., ca. 1870-1891’, with two distinct ‘J. W. Steere & Sons’ entries for Springfield, 1891 to 1894 and 1894 to 1917; the ranges are neither collapsed nor averaged. The spelling itself is dated two ways: a Stanford finding aid dates a ‘Steere & Turner’ postal cover to 1883, whereas the register and a trade-journal account use ‘Steer & Turner’ through 1891 and place the added final ‘e’ at about the 1891 reorganization; the 1883 rendering may be a normalized transcription rather than the imprint’s own spelling, so no date for the change is inferred. How that database is structured and searched is owned by a database-method page outside this group; the point here is only that consistency inside one database has to be checked rather than assumed.

The rule is to carry each position separately attributed, naming the source and its date, and not to synthesize a single answer. A record conflict yields no attribution and no date by itself; conflicting records are evidence that the question is open, and each position is carried until a record adjudicates it. Separate attribution is the disposition applied to a record conflict, not a description of the conflict: it is a recording discipline, not a finding about which record is correct. The same rule reaches any published figure that conflicts between sources of similar standing; the same evidence set applies it even to published planning intervals for organ leather, where figures from different populations are carried with their attribution, date, and population and are never averaged into an industry interval. Where a single line is unavoidable, it states only what the positions share and keeps the disagreement visible: a formation in 1898 with Hartford operation from about 1899, an incorporation in September 1919, a closing in 1971 or 1972.

A conflict ends when a record explains it, not when a reader prefers one side, and records do sometimes explain. The Rockefeller Memorial Chapel heading was settled by the owning institution’s own statement of builder, opus, and date, which is what an adjudicating record looks like: a record from a party in a position to know, addressing exactly the point in dispute. The Estey Opus 89 case shows the weaker outcome: reading the archival span as a file date explains why the two dates differ in kind, but the opus list still gives 1903 and the archival-object title still gives 1912, so the conflict is reframed and both dates remain on record.

Corroboration can split a set of pairings without settling all of them: of four Schantz rebuild-number pairings carried forward, trade sources independently corroborate Opus 2338 and Opus 2343, while Opus 2151 and Opus 2161 were found only in database search-result titles. The Organ Historical Trust of Australia attributes the Melbourne Town Hall enlargement to Schantz without an opus number and the Cleveland Orchestra archive identifies Schantz’s Severance Hall role without publishing one, so those two pairings stay open rather than being confirmed by repetition.

What does not adjudicate is a cataloger’s note, however authoritative its home: an Organ Historical Society catalog note says Francis H. Hastings had become a partner ‘by 1872’, when the corporate name became E. & G.G. Hook & Hastings, and that the firm’s assets were dispersed in 1935, while a Historic New England record, sourced to a 2006 encyclopedia article, says Hastings was promoted to partner in 1865, that ‘the company’s name was formally changed to Hook & Hastings Company in 1871’, and that the firm closed in 1936. Both notes cite published histories rather than corporate documents, ‘formally changed’ implies a filed instrument but cites none, and neither adjudicates the other. None of these examples settles anything about a reader’s own instrument; they show what kind of record can adjudicate and what kind cannot, and an attribution or a date for a specific pipe organ still rests on that instrument’s records and physical evidence.

The general ranking of evidence classes, and how records, marks, and physical inspection weigh against each other, is the evidence-hierarchy hub‘s job; this page adds only the conflict typology and its disposition. A conflict between a factory drawing and what stands in the instrument is a different evidence problem, owned by the drawings page, and patent and trademark chronology, including what a filing date, an assignee, or an abandoned application can and cannot show, belongs to the patent page. Where an owner keeps the carried positions so that a later project starts from them rather than from a single inherited number is the subject of the owner’s instrument dossier page. Opus-number systems in general, database structure and search, archive access, physical identification, component diagnosis, and builder history told as a narrative are owned elsewhere and are entered here only where two records disagree.