ApiaryActiveLive
Try: pause · settings · learn · wipe
← Community / Reading Room
RT
craft · 24 min read

Read the Mark Before Naming the Maker: A Furniture Attribution Worksheet

There is a small paper label under a drawer and three letters scored near the back of the case. A search for the letters returns a dealer's page with a…

There is a small paper label under a drawer and three letters scored near the back of the case. A search for the letters returns a dealer's page with a handsome sideboard, a plausible date and a confident maker attribution. A caretaker now has to write a label for a town display. The problem is not finding a famous name with the right initials. It is discovering whether either mark identifies the production of this object at all.

Marks are interesting because they are events. Someone applied or incised them on a particular surface at some time, for some reason. A workshop, inspector, retailer, assembler, repairer or owner might have left traces in different places. Recognizing the letters is only part of the job; evidence about which event left the mark helps distinguish the maker from someone who sold, moved, maintained or catalogued the object. This is a writing and evidence method, not an authentication or valuation service, and it does not ask people to dismantle valuable pieces.

Keep the mark where you found it

Describe the object as it stands: broad type, number and arrangement of sections, materials you can actually identify, finish if known, visible later changes, and dimensions where they can be taken without stressing it. “Brown wood with contrasting front and secondary wood inside visible drawers” is better than a species claim based entirely on surface color. A named style may help a search but should not arrive in the inventory disguised as date or authorship.

Now locate the mark relative to the part that carries it. Is it on a fixed rail, a loose drawer, an original-looking lock, an attached board, a reinforcing block, a backing card, or a loose piece kept with the object? One mark on a drawer travels if the drawer travels. A replacement lock can have its own maker. Describe the location in object terms, such as “underside of top right drawer as viewed from the front,” instead of “right side,” which changes when someone walks behind the piece.

Document the mark in context and close up where it is safe to do so. Include one photograph showing which part the mark is on, and a legible view of the surface under ordinary light. A slightly changing angle of external light can help read shallow marks if the position is safely accessible. Do not remove a panel, wet paper, rub a friable mark, apply powder, or pull a stuck drawer into the light for this exercise. A small paid or volunteer exhibition isn't improved by harming the most useful object during identification. Leave inaccessible areas marked unexamined and seek appropriate expertise when necessary.

Write a diplomatic transcription, faithful to what you see. Preserve spacing, doubled lines, punctuation and uncertainty. Put interpreted spellings in a separate field. A partial surname ending in “ER” can produce a surprising number of plausible names after you decide what you want the answer to be. Treat an unreadable area as unknown, not as permission to fill in a pattern. Where one of a set of initials might be a decorative divider, keep both readings.

Ask what the mark was doing

When an object carries more than one mark, group each by its evidence rather than ranking every mark on the same certainty scale. An institutional inventory number and a workshop brand can both be read confidently but answer entirely different questions. Give separate rows for visible form, physical placement, literal text, and candidate function. “Possible retailer label” is a candidate function; “made in this shop” is a claim that will need other support.

A full printed address and an explicit line naming a cabinetmaking business are more informative than letters alone. They still don't automatically show the label was applied to original production. A caretaker could have moved a paper label onto a replacement component. A retailer might advertise what it makes and separately sell pieces made elsewhere. A printed serial number might identify a pattern, an inventory reference or a batch, and matching it in two pieces does not by itself prove they passed through the same workbench. Separate the label's wording from the function you infer.

Context makes differences visible. The Met's catalogue record for a Bernard II van Risenburgh side table discusses both the cabinetmaker's initials and a separate guild inspection monogram. They coexist on the object; they do not supply two competing maker names. The same museum's record for a stamped commode interprets repeated initials with the help of a specific body of furniture attribution. These are well-documented examples of distinct functions and evidence, not directions for assigning any matching public mark to a particular maker.

Find an exact comparable mark if available, but demand a better match than similar type. Does the source image show the same wording, devices, arrangement and manner of marking on a plausibly related part? Is the comparative item independently attributed, or did it also get its name from a faint monogram no one can explain? An auction catalogue and an online inventory can echo the same uncertain attribution. Count them as descriptions you found, not two independent observations of the source.

Look for a hypothesis that can lose

Start with two reasonable alternatives, even when one seems obvious. Hypothesis one: a workshop label on the original case refers to the builder. Hypothesis two: a retailer marked a piece made by an unidentified supplier. If both fit the wording and known trade practice, comparing style won't automatically settle the question. You need evidence that distinguishes them, such as a documented sales pattern at the applicable time, a match to a period catalogue recorded with an independent attribution, or a clear inventory of products received from another named producer.

Useful comparison fields include overall construction, distinctive arrangement of secondary parts, dimensions, materials used in similar positions, and repeated design details. Examine features you can inspect without harm and favor combinations over single fashionable motifs. One silhouette can be popular across several workshops. A set of measurements matching a catalogue description says more if there are also unusual details in construction and decoration, but documented production patterns can still include substitutions, licensed designs and copies. Keep counterexamples in your notes rather than removing them from a visually satisfying mood board.

The Met's comparative essay about Duncan Phyfe and Charles-Honoré Lannuier explains how museum comparisons can draw on marked pieces alongside design, materials and construction, and how similar work from other craftsmen clouds an attribution. Use that as a reason for being specific, not a recipe for elevating resemblance to certainty. Also distinguish dates in biographies from dates on evidence. The date a shop occupied a location can constrain a label that lists that address, but it doesn't automatically date the wood carrying it. A known interval relevant to one component doesn't supply an exact day of manufacture.

The Met also describes an A. Kimbel and Sons étagère with a paper label, noting numbers that may relate to inventory or style. What matters for this worksheet is preserving the alternatives if the system behind a printed number is unclear. Keep the mark searchable as observed and don't upgrade the number to an authoritative model code merely because a spreadsheet column is labeled that way.

Synthetic example: the C.M. sideboard

All observations and businesses below are invented; none describe an actual dealer, chair or historical maker. A community museum receives a sideboard with a note saying it was bought on Farrier Street. A pencil mark on the inside of the lower left drawer reads “C.M.” A label on a rear board reads “Clover Mercantile, Household Furnishings, Farrier Street.” A second object in a collection photograph has rounded drawer fronts similar to the sideboard, and an online seller calls that design Clayton & Moss.

The first worksheet entry is straightforward: label lettering and address transcribed from the visible paper; rear board looks different in exposed tone from nearby boards, but the reason is unverified. The drawer entry is separate: letters visible in pencil; no explanation of their purpose found; drawer can be removed in principle but was not removed for the examination. No one records “repair indicated” just because the wood differs in appearance. Outside of the evidence notes, the provisional label says “Sideboard with Clover Mercantile retailer label; maker not identified.” If the mercantile's role as retailer were not supported by evidence, “bears a label reading Clover Mercantile” would be safer still.

A period city trade directory classifies Clover Mercantile under furniture dealers. That's evidence of the classification in one issue, not proof the firm never manufactured anything. An advertisement explicitly identifies Clover as a seller of several makers' goods; this lends more weight to the retailer hypothesis, still without naming the builder of this object. The remembered street agrees with the label, but the note came with the piece and could repeat known information. Treat it as a separate source with uncertain independence.

The Clayton & Moss claim remains possible, but not demonstrated. The seller's display does not show a mark, and the collection photograph is of a similar front only. Even if both names reduce to the same initials, that is a search lead, not corroboration. A period pattern book found later prints a comparable front under a third firm's name, an example that helps show why the silhouette is not diagnostic. “Clayton & Moss style” would now risk communicating a maker connection that no source has earned. Use descriptive words about the fronts and feet if they help the visitor actually see the object.

What might make the label stronger? A dated furnishing invoice that itemizes this sideboard with distinctive measurements, names both retailer and producer, and can reasonably be connected through ownership history would address more than style. A catalogue entry matched only to the same printed number would remain suggestive, especially if the possible meanings of that number had not been resolved. Name the missing link in your worksheet; don't tick the closest box. Even a document with a date and name needs a connection to the piece before it becomes proof.

Use a label a curator can correct without rewriting everything

Write a visible attribution sentence, a short evidence sentence, and an internal note. If the evidence supports only a literal observation, the external label can say “Stamped with the initials …” or “Bears the trade label ….” If analysis makes a particular role plausible but not established, keep the language of the uncertainty, such as “possibly retailed by …,” and briefly say why. Reserve stronger authorship claims for stronger evidence and the institution's own attribution style. Museums can have different conventions for “by,” “workshop of,” and “attributed to.” Don't use them as interchangeable decorations.

The second sentence can do positive work without smuggling in an inference. “The paper label is on the rear board; the pencil initials appear on one drawer” gives the reader something real to understand. An internal note can describe a comparison and the tests it survived, where the comparison came from, and why it did not answer the question. An accession or catalogue number names the record, not an author. Save it separately so the next editor doesn't copy it into a sentence about the maker.

Keep photographs, record links, notices about physical condition and candidate names in a folder a colleague can interpret. Store the object identifier and the photo identifier separately; several photos may document one mark, and one photo may include several. Distinguish a photograph of the actual submitted piece from a reference image used for comparison. Images of other objects all too easily move into an exhibit folder and eventually acquire a caption meant for the real object.

Handle a precise label on the wrong component

A strong inscription and weak object connection can coexist. Imagine a period lock with a clear printed name on an otherwise undated cabinet. The name might securely identify the hardware manufacture and still tell you almost nothing about when or by whom the cabinet was assembled. A documented repair might date installation to a later decade. If the lock is original to the piece, that conclusion calls for evidence about how it fits, its surroundings, its condition in relation to other documented features, and qualified interpretation where preservation is involved. It isn't established by visual neatness. A replacement part can be installed neatly.

The same is true of transferable paper. A photograph of a label still attached to a broken backing board is not the same source as a label filed in an accession envelope decades later. Both can be useful but have different chains of association. Ask whether historic photographs show the component in place, whether accession records describe it, and whether documentation associates a separate part with this specific object rather than a donation lot. If you cannot reconnect it through records, record the claimed association as a claim. Never affix separated evidence to an object as part of this research method.

Use a short role matrix. A stamp reading three letters is one row; an attached retailer label another; a contemporary carton number maybe a third. Columns might record when the mark was first documented, which part it is on, whether that part has an independently documented connection to the object, role hypotheses, and one test that distinguishes the leading roles. The last column keeps the exercise from becoming endless name matching. A strong result might leave three decent stories rather than one uncertain story: a part made by one company, a piece likely distributed by another, and a later collecting history.

Check that a public label doesn't collapse those subjects with the word “by.” “Cabinet with lock marked Fitch” leaves the object and its component legible. “Fitch cabinet” blurs them. Captions for detail photos need the same care: “The lock bears this name” is not equivalent to “Cabinetmaker's signature inside drawer.” The more visually prominent the detail, the more a reader may take the wording as the exhibit's main attribution. Put the evidence level into the sentence, not only the private worksheet. That gives visitors permission to learn something specific from a very clear name without being invited to make a larger unsupported claim.

Compare separately produced sources, not many copies of one claim

A museum record, a trade advertisement, a later dealer listing and family paperwork can all name the same maker while relying on the same attribution made decades earlier. Count the sources of the assertion rather than the number of places you can read it. An independently dated photograph that shows an existing label corroborates that the label was then present, not that everything anyone later inferred from it was correct. A modern listing copying the museum record can help locate that record but does not become a separate historical witness.

Write a tiny citation tree. With each observation, record the thing in front of the researcher and the earlier evidence it says it used. If a printed booklet quotes a dealer's tag, preserve both identities. If no source is disclosed, mark the origin unknown. In the public description, saying an object “was described as” a maker's work in an older inventory can be informative without adopting the attribution. It explains catalog history and avoids silently turning once-repeated words into a stronger claim.

A pattern book and a matching furniture surface can have a documented relationship limited to design. The Met's curatorial discussion of Phyfe and Lannuier discusses marked and attributed works in the broader context of dimensions, materials, design and construction. A recognized decorative pattern can identify what artisans may have seen or used, while many shops could execute comparable forms. Write “features are comparable to” before going any further, and say what would have to distinguish the candidate workshop from others. The best negative comparison is a confidently separate workshop known to have used the same visual feature.

Consider a printed trade label appearing in one book and a photograph of a similar label published years later. If the image belongs to the same furniture item or was reproduced from the first source's photograph, their agreement is not a replication across objects. If two independent objects have the same mark printed in a distinct configuration, the comparison might strengthen the existence and consistency of a marking practice. But the role of the mark and the date it was applied are still separate. Preserve exactly which claim got stronger and which is waiting on another kind of record.

A second worked example: two desks and a borrowed maker's stamp

The furniture, museum, marks, shops and records in this example are synthetic. A local exhibition designer finds two small desks with similar slanted fronts. The first desk bears K.V. incised on a metal hinge and a paper label under a loose writing tray. The second, from a separately documented acquisition, has the same stamp on two hinges but no paper label. Someone says that both were made by the hypothetical firm Kent and Vale, which appears in a trade listing for cabinet work. A clear photograph of the hinge mark makes the proposal feel more solid than the evidence warrants.

The worksheet starts by recording the marks as observed. Both K.V. impressions are on hinge metal, not on a fixed wooden outer surface. The first desk's paper label names an outfitter and includes a job reference, but its recorded attachment is to a tray that can be moved. No direct comparison shows that the tray belonged to the desk at original assembly. The case and tray have separate research notes. The inspector does not remove hinges or any other component to identify them. Visible observations and documented history are the basis for the exercise.

An independent fictional hardware catalogue matches the stamping pattern to a supplier of fittings, not to the cabinetmaker previously proposed from that acronym. It contains a similar hinge but allows for several batches. This can make the hardware interpretation more persuasive without proving precisely when the hinges were installed. On the second desk, the hinges and visible surface have different histories unknown to researchers; installer, fitter and builder need not be the same. The two desks may both contain parts supplied by that firm. They do not now have a verified common overall maker.

Then a small archive finds a correctly linked shipping list referring to one of the desks by a distinctive combination of dimensions and an otherwise unusual plate. The list attributes delivery to the outfitter named on the tray and names a different workshop as builder. The identifiers and comparisons support candidate identity in this invented case, but the research team still checks whether this list and the physical object belong together before revising the label. The second desk lacks the same reference. Its maker should not be carried along because both share hinge markings. A case-specific inference is not a group rule.

The public labels can reflect this difference. The first can name the outfitter and workshop when the record linkage is verified, make the role distinction clear, and mention the hardware maker only if warranted. The second simply tells readers that the hinge carries a mark associated with the hardware supplier, with the builder not established. These labels are not equally glamorous, and that's fine. The difference teaches a more interesting history of how sellers, parts makers and workshops interacted. It also preserves the unresolved question for a later researcher without allowing one good attribution to certify every similar object.

The lesson does not depend on finding a famous maker. A useful finding might be that paper label and stamp identify different functions, or that the two desk fronts should not be treated as evidence of an exclusive workshop design. Research earned that narrower description by checking how each mark related to the item. If the curator later receives a competing contemporary source, the worksheet tells them exactly what must be reconciled rather than forcing them to defend a confident caption written without a documented chain.

Choose comparative photographs that preserve uncertainty

A clean front view is excellent for an exhibition and often weak for an attribution. Distinctive details can live in interior construction, surface attachment, or the relationship between a label and a constituent part. A photographic comparison should be cared for like any other source: record which object and detail appear, when it was made if known, what visible dimensions or orientation it includes, and where you got permission to use it if needed. A photograph of a comparable advertised object cannot stand in as a photograph of your own. A cropped plate that omits the edge of the object may not let a reader identify which part it shows.

Equip the viewer to locate the mark without overclaiming what the photograph reveals. A wide view can orient from front, rear or distinctive hardware. A matching detail can show the letters or outline. If access is poor and key details blur, state that instead of reshaping the contrast until a hoped-for letter appears. Treat changes in brightness or tonal range as processing choices. Keep a stable reference to the original as allowed. Don't submit a dramatically modified view to a specialist as though it were the only direct observation. They need to know whether uncertainty was in the original surface, camera, or edited copy.

The most useful comparison pictures may be the imperfect ones that answer the right question. A documented cabinet with an identically worded trade advertisement inside a replaced drawer could be an example that separates retailer evidence from manufacturing rather than proves the specific drawer belongs to a workshop. Note whether the relationship was determined from curatorial examination or only observed in a photo. If you cannot safely inspect a part yourself, do not ask an untrained person to move, open or alter a valuable or unstable item for the shot. A precise documentation gap is a legitimate next action to give the collection's responsible staff.

Put date ranges on the specific claim that earned them

A guessed year tends to spread through an exhibit. It starts as the date of a particular directory entry, then becomes the year a label was applied, and by publication the year the furniture was finished. Preserve what each piece of evidence actually dates. A printed catalogue can establish when a seller advertised a model. A dated photograph shows an object in that view at a certain time if the photograph and its date are verified. A later repair receipt may establish replacement activity. None automatically supplies the absolute age of every piece in the photograph or furnishing.

A mark bearing a named address can constrain an interpretation when the history of the firm and that address is legitimately documented, but ask what else must be true. A label might have been used after its location was printed, be saved to attach later, or refer to a dealer not the maker. Treat any temporal inference as conditional on that relationship. When the condition fails, update the relevant inference rather than replacing established dates with an unsupported one. A chronology with three different evidence strengths is manageable; a caption pretending they all carry the same strength is difficult to correct.

For a short exhibition label, choose the narrowest supported date statement and document a fuller note internally. A fixture can have a dated component but a broader period of appearance. “Fitted with hardware from a supplier active during this interval” is a different statement from “made in this interval by the supplier.” For some research purposes a rough date of use is enough. For others, the order in which parts and labels were introduced is the actual subject. Don't collapse these stories so an attractive first sentence can bear more than any source supports.

Preserve how a change in attribution was made

If a catalogued attribution changes, keep an audit note with former wording, relevant evidence, decision date and the revised description. A researcher encountering an older catalogue should be able to understand how it relates to the updated label without having to subtract one attribution from another. If older language had a different role, classify it, rather than treating it as either equally accepted or deliberately suppressed. Preferred language can express role and uncertainty in one sentence while internal notes carry the longer explanation.

An evidence correction may apply only to part of a record. Updating a lock supplier does not necessarily affect the item description or its display in the dates section. Moving an image from one comparative item to another might change which captions cite it without altering the object's physical inventory. Give each correction a scope and a verification step. Reread the labels that rely on that evidence, including images and separate educational material. Correcting the museum database is not proof that every derivative output changed. Track verified changes and leave other outputs unresolved until checked.

A reader-facing label should not reveal private documentation or unsupported rumors while explaining the reasoning. Often two lines are enough: the name as literally marked and the better-supported role or uncertainty. Internal evidence can establish the provenance of the observation and why the expert changed their interpretation. That internal distinction may be useful if another staff member later encounters the object in an older collection. Clearing a bad association can be an achievement even when the finished public text becomes shorter.

Read a written role without confusing it with a maker's signature

Some labels combine a shop name, advertised trade, place, and general call to purchase. Even a direct statement that a firm sold cabinets does not establish that this firm manufactured a particular case. First record the exact words you can see and the placement. Then ask whether the mark names a maker, retailer, repair service, administrative owner, a part's manufacturer or a function not yet known. If the handoff needs a concise inventory field, keep that field provisional and retain the evidence explaining the role. An unclear role should not be hidden behind an authoritative-looking maker drop-down.

A team can compare exact label wording across verified examples. Does a similar shop name have the same address and layout? Are the objects from an independently documented time and manufacturing context? Is there a reason to believe two photographs show different surfaces of one shared object rather than two independently labeled examples? These questions matter more to an attribution than the number of matching image results. If the word indicating the trade is partly missing, don't attach a specific function unless other appropriate sources support it. A comparison may connect a print shop, a merchant, or an earlier owner without proving the physical maker of the furniture.

One useful case assignment asks a volunteer to find evidence against the lone favorite interpretation. If the favorite maker is inferred from the retailer's name, look for independently documented variety in that retailer's suppliers. If a repeatedly cited mark is found on hardware, locate securely identified comparable hardware before expanding the clue to cabinet production. If all comparison images originated in a single business archive, discover what that archive's coverage limits were. A team with a generous interpretation of negative evidence should still distinguish “could not confirm retailer as maker” from “confirmed retailer never made furniture.” The second is a broader claim and isn't necessary to correct a premature attribution.

Track fabrication, repair and collecting events side by side

A short event ledger can preserve relationships the final label doesn't need to explain. In fabrication, a workshop and parts supplier might be relevant. Later remodeling or repair can introduce a marked component. Acquisition can introduce a private inventory notation or institutional number. A relocation can move a label or require different wrapping that later looks like part of the object. These are kinds of events to consider, not a chronology to invent. Include only the dated or reasonably attributed observations you actually hold and mark what remains a possibility.

In a fictional inventory test, suppose one document describes a top as newly refitted but gives no firm name. A photograph shows a manufacturer's stamp on a lock in the top. A family's collecting note attributes the entire piece to that manufacturer, citing the visible stamp. The repair report may strengthen the reason to consider a later installation, though it does not prove the lock arrived with the repair. The photograph links the mark to a visible component, not a year of insertion. The family note documents a genuine attribution tradition that may derive from that same visible mark. A responsible label can explain that connection without declaring the tradition confirmed as workshop history.

A component other than the one attracting attention may supply better evidence. A verified acquisition record could pose a question about the case's decorative structure independent of the fitting's date. A well-described comparable piece might be especially informative about an uncommon combination of features but won't necessarily tell you when the family repaired its version. Keep competing events within their limits. This prevents one promising find from inadvertently revising more of the object history than it supports. In the final handoff, tell people which event is established, which would be supported by a documented contributor, and which work should be left to the next appropriate specialist.

A review meeting can save time by assigning one question to each unresolved link. “Confirm what the stamped name meant in this parts catalogue” isn't the same request as “determine the case maker.” If the first can be resolved without handling the object, start there. If the second requires examination beyond staff capability, preserve the observations and seek authorized expertise instead of staging an unsafe experiment. The historic piece is not improved by forcing a decision before its evidence is available. A well-bounded unresolved attribution can be the appropriate end of an otherwise productive project.

Make the next research hour count

Prioritize checks that could eliminate a tempting attribution. If the guess depends on a retailer serving only its own workshop, search for the company's explicit evidence of reselling before assembling another gallery of similar surfaces. If two letters on a lock are the primary clue, ask whether a period maker mark register or comparable object identifies lock manufacturers rather than cabinetmakers. If a claimed construction trait appears across workshops, record that and stop assigning weight to each additional example of that trait. Resources vary by place and period; don't treat the first plausible directory hit as a complete register of every woodworker.

A simple status vocabulary is enough: observed mark, role supported, attributed from independent evidence, or not established. Apply it to each mark and conclusion individually. “Role supported” can be a satisfying result. A piece traced to a retailer tells a story about who handled and sold furniture in a place, even where its exact workshop remains unknown. Protecting that narrower discovery often produces a better exhibition than turning every furniture label into a list of famous makers.

The most useful worksheet question is “What would make us change this label?” If the answer is a certain dated document, an independently attributed comparison, or a look at a surface by a qualified person, it's a real question, and the next researcher can act on it. If the answer is merely finding another picture with similar legs, the evidence hasn't yet given an object-wide attribution. Keep the mark, record the inference, and give the furniture a description that won't have to be unlearned.

Editorial note: Drafted with AI assistance; the C.M. sideboard and related evidence are synthetic.

Frequently asked
What is Read the Mark Before Naming the Maker: A Furniture Attribution Worksheet about?
There is a small paper label under a drawer and three letters scored near the back of the case. A search for the letters returns a dealer's page with a…
What should you know about keep the mark where you found it?
Describe the object as it stands: broad type, number and arrangement of sections, materials you can actually identify, finish if known, visible later changes, and dimensions where they can be taken without stressing it. “Brown wood with contrasting front and secondary wood inside visible drawers” is better than a…
What should you know about ask what the mark was doing?
When an object carries more than one mark, group each by its evidence rather than ranking every mark on the same certainty scale. An institutional inventory number and a workshop brand can both be read confidently but answer entirely different questions. Give separate rows for visible form, physical placement,…
What should you know about look for a hypothesis that can lose?
Start with two reasonable alternatives, even when one seems obvious. Hypothesis one: a workshop label on the original case refers to the builder. Hypothesis two: a retailer marked a piece made by an unidentified supplier. If both fit the wording and known trade practice, comparing style won't automatically settle the…
What should you know about synthetic example: the C.M. sideboard?
All observations and businesses below are invented; none describe an actual dealer, chair or historical maker. A community museum receives a sideboard with a note saying it was bought on Farrier Street. A pencil mark on the inside of the lower left drawer reads “C.M.” A label on a rear board reads “Clover Mercantile,…
References & sources
  1. Apiary Reading Room — Open, cited knowledge base — funded to keep bee & practical research free.
From the Apiary Reading Room. Opinion & editorial — not financial advice. We don't overclaim.
More from the Reading Room