Gray graphics madness

I am a sucker for a good graphic, and the recent Rand report into illicit antiquities has a bunch of good ones, in particular a quadrant plot used for exploring the structure of the antiquities market (Rand 2020: figures S.1, 4.10, 4.11, 4.12, 4.13). This graphic offers a means of looking at market structures generally, using axes of visibility (covert to overt) and legality (legal to illegal) to construct a conceptual transaction space. Two of the quadrants (covert-legal and open-illegal) are considered to demarcate gray market space, with the antiquities market placed firmly towards the overt end of the visibility axis and straddling the divide between legal and illegal on the legality axis (Figure 1).

Figure 1 (Rand 2020: figure 4.1)

Once I had managed to overcome my initial excitement at seeing the graphic, I became concerned about how it portrays the gray market. The idea of a ‘gray’ trade or market was introduced into criminological discussions of antiquities trading about 15 years ago. There was some confusion about the concept until it was systematized by Simon Mackenzie and Donna Yates (2019), who identified three possible usages:

  1. Mixed streams of supply. Legally-obtained and illegally-obtained antiquities sold on the open market cannot be distinguished because of undisclosed provenance. An antiquity with an undisclosed provenance is regarded as gray, possibly on the market legally (white) or illegally (black).
  2. Changing status of individual objects passing through trafficking networks. An illegally-obtained antiquity can be cleaned through time and space as it moves through jurisdictions up the trading chain for sale on the open market. So although its sale is technically legal, the antiquity might still be considered illicit, having been subject to an illegal transaction at some point in its trading history. Legally white but ethically black – gray.
  3. Neutralization and the graying of the moral psychological processes of engagement. Actors justify or neutralize their engagements with illegally-obtained antiquities by appealing to a greater common good. Thus their engagement can be considered somewhere between good and bad – gray.  

The Rand report recognizes Mackenzie & Yates (1) when it says:

Although illicit drugs, for example, are primarily sold on the black market because of their near universal illegality and strong law enforcement action against them, illicit antiquities can be advertised and sold through more-visible channels because it is often difficult to prove an individual item’s illegality (Rand 2020: xiii).

Unfortunately, the report doesn’t use this definition for defining its own gray market quadrants, deciding instead that:

… illegal items are sold relatively openly and, conversely, legal transactions may be conducted in private. These two cases are classified as gray market transactions (Rand 2020: 63).

While Mackenzie and Yates focused upon the status of the object (1 & 2) or actor (3), the Rand report has looked instead at market circumstances:

  1. Illegal transactions conducted overtly
  2. Legal transactions conducted covertly

Thus the Rand report has defined and portrayed the ‘gray market’ differently to most established usage in the literature dealing with the antiquities trade. This is not necessarily a problem, as the report’s definition might better suit its own purposes, but it does threaten to cause confusion if the graphic is used unwittingly by others to depict what is usually understood to be the gray antiquities market, when it does nothing of the sort (and doesn’t claim to). Having said that, it is not clear to me anyway why the two quadrants are marked gray. Their grayness does not feature in the associated analyses of market structure and appears superfluous.

Prompted by this gray mood to look more closely at the graphic, I began to think there are some deeper problems. It shows the antiquities market straddling the line between legal and illegal, though firmly placed on the overt side of the plot. While it is true that many antiquities are offered for sale openly, either at traditional bricks-and-mortar establishments such as Sotheby’s and Christie’s or on electronic platforms such as eBay, many more transactions, legal or illegal and often high-value, occur privately – in other words covertly. So, for an accurate representation of the antiquities market, the defining circle would need to be expanded outwards into the covert side of the plot (Figure 2). That would have the disadvantage of suggesting visually but inaccurately that the antiquities trade is much larger than the drugs trade, which is clearly not the case, and something the Rand report is rightly concerned not to imply. 

Figure 2

Rand figure 4.11 (Figure 1 here) is used to argue that antiquities are not sold on the Darknet because they are not generally sold covertly on the black market. Unfortunately, they are. They are routinely transacted covertly and illegally. The 2010 Hobby Lobby private (covert) purchase in the United Arab Emirates of looted Iraqi antiquities offers just one example. There are many more. Most of the source trade in countries of origin is covert and illegal. It is true that antiquities are not traded on the Darknet, but it is not because they are not traded covertly and illegally as the graphic suggests.

Perhaps I am being unfair. In its introduction, explaining Figure S1, the report states:

We first conceptualized the antiquities market in relation to other illicit markets by considering it in terms of two characteristics: the openness versus secrecy of the transactions and their relative legality or illegality. Viewed in these terms, the differences between the illicit antiquities trade and drug or weapon trafficking is more clearly evident (Rand 2020: xiii).

This introductory explanation gives the clear impression that the graphic is meant to describe the antiquities market in its entirety, which is what I initially thought. Later, however, when the graphic is used to analyse market structure, as in Figure 1 here, the report states it is a “structural analysis of online markets”, presumably not including the bricks-and-mortar trade. Almost by definition, antiquities sold online are sold overtly, and perhaps there is no need to sell them covertly, as the report states (page xx). But that doesn’t mean that no antiquities are sold covertly. But looking more closely still, there is also confusion over what market exactly the graphic is meant to be depicting – source or destination? In countries such as Syria and Iraq, antiquities sales are illegal and mostly covert. Even when antiquities are offered for sale on platforms such as Facebook, it is in closed groups open only to vetted access and further limited by the use of a local language (in this case Arabic), and as the report notes (page xx) actual transactions take place covertly, often using encrypted messenger apps such as WhatsApp. An alternative characterisation would place this source market in the covert-illegal quadrant (Figure 3).

Figure 3

On the destination market, Facebook isn’t a major platform for marketing antiquities. As the report states, ‘auctions and online storefronts dominate’, where illegally-obtained antiquities can be sold on the open market. On the graphic, this market would be placed in the overt-illegal quadrant (Figure 4).

Figure 4

The point to be made here is that when travelling from source to destination, antiquities transition across quadrants because of the gray trade, in the sense of Mackenzie & Yates (1), whereby provenances are suppressed and invented so that antiquities are not identifiably illegal by the time they appear on the open market (Figure 5).

Figure 5

Another cause of initial confusion for me was how ‘legality’ could be a continuous variable – surely it is a binary variable: a transaction can be legal or illegal, but not ‘just about legal’, or ‘tending to the illegal’ etc. But having sourced the graphic back as far as 2006 to a paper by H. Brinton Milward and Jörg Raab (2006: 335, figure 1; reproduced here as Figure 6), I think I understand it better. Their idea was to characterise dark and bright networks in terms of their overall legality and visibility, and the Rand report seems to be doing likewise for markets instead of networks, or markets conceived as networks.

Figure 6 (Milward and Raab 2006: 335, figure 1)

That observation does raise an interesting question, however. Could the transaction space be used to plot objects instead of networks? Could it be used to plot the progress of a single antiquity, thereby showing in detail how the gray trade operates? On Figure 7, I have plotted the known biography of the Gilgamesh Dream tablet using six real and one hypothetical transactions.

T1. Sometime before 2001, the tablet was looted in Iraq and bought in Jordan covertly by a dealer who must have recognised it to have been on the market illegally.

T2. In 2003, a US dealer visited London and bought the tablet covertly from the Jordanian dealer. The US dealer must also have recognised the tablet to have been obtained illegally, or at least strongly suspected it to be the case, as he invented a fake provenance in the shape of a letter stating that the tablet had been sold in San Francisco in 1981.

T3. In 2007, the US dealer sold the tablet covertly to another US dealer, along with the newly invented provenance to show that it was ‘clean’. At that point, the tablet was grayed, with its true, illegal provenance suppressed – an example of Mackenzie & Yates (1).

T4. Later in 2007, still in the United States, the second US dealer offered the tablet for sale overtly in a sales catalogue, accompanied by the new provenance.

T5. In 2014, the then owner of the tablet sold it to Hobby Lobby through a private (covert) sale arranged by Christie’s auction house in London. Depending upon what had happened to the tablet between 2007 and 2014, and what jurisdictions it had passed through, it might have been on the market in London legally, through with a fraudulent provenance – an example of Mackenzie & Yates (2). That remains to be determined in court, and that is the problem of the gray market. It is difficult if not impossible to ascertain the exact legal status of an object offered for sale.

T6. Hobby Lobby put the tablet on display in the Museum of the Bible, but subsequently recognising that it had probably been taken illegally from Iraq, agreed to return it.

? T7. In the unlikely event that Iraq decides to sell the tablet, the transaction would be open and legal.

Figure 7

This object biography shows how the illegally-obtained tablet was grayed through the invention of a fake provenance and also possibly by its (hypothetical) movement through jurisdictions and the passage of time. Time is a crucial element here. As time passes, and illegal actions diminish into the distancing past, gray antiquities slowly fade to white.

Figure 8

Another problem with the original graphic is that it doesn’t take adequate account of provenance. Provenance is a defining feature of the antiquities trade in a way that it isn’t for the drugs trade, as the Gilgamesh Dream tablet biography shows. Unfortunately, provenance adds another variable that cannot be included in the two-dimensional transaction space. But if the space is constructed differently using visibility and provenance as its defining axes, binary legality can then be indicated using symbol colour. In Figure 8, I have indicated illegal objects with a known provenance as black, legal objects with a known provenance as white, and objects of undisclosed provenance as gray. Gray objects are transacted covertly and overtly. White (legal) objects are also transacted covertly (privately) and overtly. Black objects, that is to say objects known to be on the market illegally, can only be sold covertly (unless by someone with a prison-wish). These might seem trivial observations, but they demonstrate a fundamental truth of the antiquities trade, which is that it is not possible to talk about a black market, at least not in this transaction space. The market can be invisible, in that legal (white) or illegal (black) antiquities can be sold covertly; or gray in the sense of Mackenzie & Yates (1) in that legal and illegal antiquities can be sold (overtly and covertly) with an undisclosed or false provenance; or white, in that legal (white) antiquities can be sold overtly. The policy challenge is to force a white market by improving provenance provision, as shown in Figure 9.

Figure 9

Although I was initially excited by the Rand graphic, I found much to disagree with, yet disagreement in itself has been a stimulating exercise. The Rand report has done us all a great service by suggesting how such simple graphics might be used profitably to explore the structure of the antiquities market and its structural relations with other markets. I look forward to seeing more.

(Notice I didn’t call this post ‘fifty shades of gray’. I am better than that).


Mackenzie, Simon and Donna Yates (2016). What is grey about the ‘grey market’ in antiquities, in J. Beckert and M. Dewey (eds), The Architecture of Illegal Markets: Towards an Economic Sociology of Illegality in the Economy. Oxford: Oxford University Press.

Milward, H. Brinton and Jörg Raab (2006). Dark networks as organizational problems: elements of a theory, International Public Management Journal 9: 333-360.

Rand (2020). Sargent, Matthew, James V. Marrone, Alexandra Evans, Bilyana Lilly, Erik Nemeth, Stephen Dalzell, Tracking and Disrupting the Illicit Antiquities Trade with Open-source Data. Santa Monica: Rand.