In 2003, the Bolton Museum in Lancashire paid £440,000 for a small alabaster figurine identified as an Egyptian princess from the Amarna period, circa 1350 BC. The piece had passed through Christie's with a plausible attribution and the documentation that surrounded it appeared to support an ancient origin. Three years later, police investigating an unrelated matter searched the home of a retired gardener named Shaun Greenhalgh in Bolton. In a shed behind the house, they found the tools and materials used to produce the figurine. Greenhalgh had made it himself, in the same shed, using materials available at any art supply shop and a reference book from the local library. He was convicted of fraud in 2007.
The Bolton case is not unusual in its outcome -- the fake was eventually identified and the forger convicted -- but it is instructive in how long it took, what it took to expose, and what the exposure revealed about the mechanisms by which fakes enter the market. The Greenhalgh family had been supplying fakes to the auction market and to dealers for seventeen years before their operation was uncovered. They had sold pieces to Christie's, to Sotheby's, to the British Museum, and to private collectors. The total value of the fakes they placed was estimated at around £850,000.
How fakes enter the antiques market, and what eventually catches them, is a question with practical implications for every collector. The answer involves the intersection of expertise, incentives, and the specific vulnerabilities that the market's structure creates.
Not all objects that turn out not to be what they were sold as are fakes in the same sense. The word covers a range of phenomena that have different implications for how they enter the market, how they are detected, and what responsibility attaches to the various parties who handled them.
At one end of the spectrum is the period reproduction: an object made in the style of an earlier period, by craftspeople of the later period, without necessarily any intent to deceive in the original context. Victorian reproductions of Georgian furniture were made in large quantities for a market that wanted the look of the earlier style at a lower cost. These objects are now genuinely old, genuinely well-made, and in many cases difficult to distinguish from the originals they imitated. They enter the market not as deliberate fakes but as objects whose identity has been lost or confused over time, and their detection requires the same kind of specialist knowledge as the detection of a deliberate forgery.
At the other end is the deliberate forgery: an object made with the specific intent to deceive a buyer into believing it is something it is not. The Greenhalgh family's production falls into this category. So does the work of the Dutch forger Han van Meegeren, who produced Vermeer paintings that fooled the market and institutional buyers for years in the mid-twentieth century. Deliberate forgeries are constructed around the specific vulnerabilities of the market they target -- the gaps in documentation, the limits of available testing, the psychological tendency of experts to see what they expect to see in an object that looks right.
Between these two ends sit the upgraded piece, the composite, and the misattribution. An upgraded piece is a genuine object that has been altered to appear rarer or more valuable -- a piece of silver with a maker's mark added, a painting with an attribution added in a later hand, a piece of furniture with later elements replaced with period ones that bring it closer to a desirable type. A composite is an object assembled from parts of different genuine pieces to create something that appears whole and coherent. A misattribution is a genuine object correctly described in terms of its physical characteristics but wrongly attributed to a specific maker, workshop, or period.
The most important tool available to a forger is not technical skill but documentary skill. A physically convincing fake that lacks a plausible history is much easier to detect than one accompanied by documentation that places it in the right hands at the right times. The construction of false provenance is a significant part of the effort that goes into the most sophisticated fakes, and it exploits specific vulnerabilities in how the market handles historical documentation.
False provenance typically works by attaching a fake object to a genuine documentary history. A sale catalogue from the early twentieth century that describes an object in terms that could apply to the fake in question can be used to suggest that the two are the same object. An old photograph that appears to show the piece in a credible interior creates a visual record of its apparent age. A handwritten label or letter that refers to an object and can be associated with the fake places it in a documented context.
The Greenhalgh family used exactly these methods. Their pieces came with labels, letters, and documentation that appeared to give them a history. The documentation was examined along with the objects and contributed to the assessments that led to sales. In retrospect, the documentation showed signs that specialist examination would have identified as inconsistent with the claimed periods -- the wrong typefaces, anachronistic terminology, paper that did not match the supposed date. But these inconsistencies were not identified at the time, in part because the documentary evidence was examined with the same expectation of authenticity that was applied to the objects.
The detection of fakes in the antiques market relies primarily on specialist expertise: the accumulated knowledge of people who have handled large quantities of genuine material and developed the pattern recognition to identify when something does not conform to expectations. This expertise is genuine and effective, but it has structural limits that sophisticated forgers learn to exploit.
The first limit is that expertise is specialised. A specialist in English silver of the eighteenth century may have exceptional knowledge of hallmarks, construction methods, and the stylistic characteristics of specific workshops, and will recognise a fake in that category that would deceive a generalist. But the same specialist may be less reliable when asked to assess a piece that falls at the edge of their knowledge -- a piece from a related but distinct tradition, or from a period adjacent to their specialism. Forgers who understand the geography of specialist expertise can sometimes find the gaps.
The second limit is confirmation bias. An object that arrives with compelling provenance, through a reputable channel, with an attribution that fits the documented history, invites assessment that confirms what the documentation suggests rather than testing whether the documentation is correct. This is not a failure of individual experts but a structural feature of how assessment works: experts who see thousands of objects develop efficient pattern-matching processes that correctly identify the vast majority of what they assess, and the occasional sophisticated fake succeeds by satisfying those patterns rather than triggering them.
The third limit is the social structure of the market. Auction houses and major dealers have reputations to protect and are motivated to catch fakes, but they are also motivated to sell material and to maintain relationships with consignors. A consignor who brings significant material to a house expects a certain handling, and an assessment that delays or prevents a sale has costs that a passing assessment does not. This is not to suggest that commercial pressure systematically overrides professional judgment -- in most cases it does not -- but it creates an environment in which the marginal fake, the one that is not obviously wrong, may receive less scrutiny than it would in a pure assessment context.
The development of scientific methods for testing the age and composition of materials has significantly changed the landscape of fake detection in the antiques market. Thermoluminescence dating, which measures the accumulated radiation dose in ceramic materials, can establish when a piece of pottery was last fired and does not depend on stylistic judgment. X-ray fluorescence analysis can identify the elemental composition of metals and compare them against the known compositions of historical alloys. Carbon-14 dating can establish the age of organic materials within a range of years. Dendrochronology can date wooden panels by the pattern of tree rings visible in the wood.
These methods are not universally applicable. They require access to the object, they have costs, and they have limits -- thermoluminescence testing can be defeated by exposing a newly made ceramic to artificial radiation sources, carbon dating is less precise for objects from the past few centuries than for older material, and X-ray fluorescence cannot determine when a piece was made, only what it is made of. But they provide a form of evidence that does not depend on the subjective assessment of experts, and they have caught fakes that would have been extremely difficult to detect by visual examination alone.
The use of scientific testing in the market is uneven. Major auction houses routinely use testing for certain categories -- archaeological material, ancient ceramics, objects where scientific evidence is directly applicable. For other categories, testing is used selectively, when an assessment raises questions that visual examination cannot resolve. Private buyers who want scientific reassurance about a significant purchase can commission independent testing, and the results are sometimes used in due diligence processes that precede major acquisitions.
The category of period reproduction creates a specific problem that is distinct from the deliberate forgery. An object made in 1880 to look like a piece from 1680 is two hundred years old when examined today. It may show genuine age, genuine wear, and the physical characteristics associated with old material. The question is not whether it is old but whether it is old in the way that was represented -- whether the age it shows is the age it was made to simulate, or the age it actually has.
This question is most acute in categories where period reproductions were made in large quantities and with high skill. Victorian cabinet-makers who reproduced eighteenth-century English furniture were working in a tradition of craft that had not changed dramatically since the originals were made, using similar tools and similar materials, and the results can be difficult to distinguish from genuine period pieces without specialist examination. The same applies to reproductions of medieval metalwork made for the Victorian collecting market, to Orientalist reproductions of Islamic ceramics and bronzes, and to the vast production of reproductions in various categories that accompanied the collecting booms of the nineteenth and early twentieth centuries.
For a collector, the distinction between a period reproduction and a genuine period piece is significant in terms of value but may be less significant in terms of the object's interest and quality. A skilled reproduction of an eighteenth-century piece made in the Victorian period is itself an historical object with its own qualities. The problem arises when it is represented as something it is not, which may happen through deliberate misrepresentation or through the accumulated confusion of a long and undocumented history.
The distribution of fakes in the antiques market is not random. They concentrate in categories where the potential reward is high relative to the difficulty of production, where the relevant expertise is difficult to access, and where the documentation that would expose them is incomplete or unavailable.
High-value categories with limited supply are most attractive to forgers: ancient art and archaeology, where genuine pieces are rare and scientific testing is the primary check on authenticity; furniture of the most desirable periods and makers; silver with specific hallmarks; and art pottery from the most sought-after kilns and studios. These are also the categories where the collector who pays a significant premium for an object has the most to lose from a misidentification.
Categories where scientific testing is directly applicable are in some respects better protected than categories where assessment relies entirely on connoisseurship. An ancient ceramic can be dated by thermoluminescence. A piece of furniture cannot be dated by the same method, and the detection of period reproductions in furniture relies on close examination of construction details, tool marks, the way surfaces have aged, and the consistency of the piece with the claimed period -- all of which are matters of specialist judgment rather than scientific measurement.
Among the most common forms of misrepresentation in the antiques market are not outright fakes but objects that are genuine in part: pieces assembled from components of different periods, or genuine pieces that have been enhanced to make them appear rarer or more desirable than they are.
A piece of furniture assembled from parts of two or more period pieces may be entirely genuine in its components while being inauthentic as an object -- it never existed as a single piece in its current form and the history implied by its appearance is fictional. The detection of composite furniture requires close examination of the joints, the consistency of the timber, the relationship between the different elements, and the evidence of how and when the various parts were joined. An experienced specialist in period furniture can often identify a composite by the inconsistencies in these details, but the examination requires time and access that is not always available in a busy sale context.
Upgraded pieces present a similar problem. A piece of silver with a later-added hallmark is partly genuine -- the silver exists and may be period -- but misrepresented in terms of its maker or period of manufacture. The added mark can sometimes be identified by close examination of the surrounding area: the engraving style may differ slightly from the surrounding surface, the depth of the mark may be inconsistent with wear, or the design of the mark may not match the documented output of the claimed maker. But these are details that require specialist knowledge to identify, and a mark that has been added with care can be difficult to detect without scientific examination of the metal surface.
The mechanisms by which fakes are eventually identified vary by category and by the sophistication of the fake in question. For the majority of cases, the initial identification comes from specialist visual assessment -- a dealer or specialist who has handled enough genuine material recognises something that does not fit the expected pattern. The surface is not right, the construction is not consistent with the claimed period, the style includes anachronistic details that the maker or period would not have used.
For the more sophisticated fakes, the initial visual assessment may not raise flags, and identification comes from scientific testing prompted by other circumstances: a collector who commissions testing as part of due diligence, a researcher who examines a piece in connection with a catalogue raisonné, a conservation examination that reveals unexpected materials. The Greenhalgh fakes were not identified by specialist assessment of the objects themselves -- they were identified because police searching the property for unrelated reasons found the workshop and the evidence of production.
This is a significant point: the most sophisticated fakes may remain undetected until circumstances beyond the market's own assessment processes expose them. A collector who relies solely on the assessment that accompanied a purchase, without commissioning independent verification for significant acquisitions, is relying on a system that has demonstrable limits. Due diligence that includes specialist examination and, where applicable, scientific testing is the most reliable protection available, and it is most important precisely for the categories and price levels where fakes are most likely to be found.
The prevalence of fakes in the antiques market does not mean that the market is fundamentally unreliable, and most objects sold through reputable channels are what they are represented to be. But the cases where they are not are instructive about what protects the collector and what does not.
Reputable provenance helps but is not conclusive. The Greenhalgh fakes passed through Christie's. Objects with impeccable auction records have been subsequently identified as inauthentic. Provenance provides evidence that should be weighted in an assessment, not a guarantee that removes the need for one.
Specialist assessment by someone with direct expertise in the specific category, who has examined large quantities of genuine material of the relevant type, is the most reliable form of protection available. The assessment of a generalist, however experienced and reputable, is less reliable than the assessment of a specialist for a specific category. Identifying and engaging the right specialist for a significant purchase is part of the due diligence process.
Scientific testing, where applicable, provides a form of evidence that does not depend on subjective judgment and that sophisticated forgers have more difficulty defeating than visual assessment. For categories where testing is applicable -- ancient ceramics, organic materials that can be carbon dated, metals whose composition can be compared against historical standards -- commissioning testing for significant acquisitions is a form of insurance that the cost of the test usually justifies.
Understanding the categories where fakes concentrate, and where the risk is highest, is itself a form of protection. A collector who understands that certain categories carry higher fake risk than others, and who brings appropriate scrutiny to purchases in those categories, is better protected than one who applies uniform levels of attention regardless of risk profile. The market is not uniformly dangerous, and the collector who understands its geography is better positioned than one who does not.
Orpheus Alerts combines AI monitoring with a specialist network of human sources (dealers, art advisors, estate contacts, and private sellers) to track what you're looking for across auction houses, galleries, private sales, and channels that never appear in a standard search.
Tell us what you want. We cover the sources where it is most likely to surface. When we find a credible match, you receive a direct alert with a link. No noise. No irrelevant suggestions. Just the thing you've been looking for, when it appears. Orpheus Alerts — When you know what you want.
