By Giorgio Torraca

A scientist in conservation may get involved in preserving material of historic or artistic value in various branches of the profession and at various levels—working directly with the material itself to doing basic research. A scientist can work in analysis for archaeometry (how was an object made? when? where? was it modified later?) or for conservation (is it deteriorating? why? how fast?). He can work in technology for the restoration of anything from small artifacts to large monuments (e.g., cleaning and consolidation processes, immediate and long-range effects) or in technology for protection (e.g., modification of the environment, protective layers, maintenance procedures).

In the past, the choice of the scientist was mainly dictated by chance—a friend asking, "Would you take a look at our problem?" Then he might remain trapped for quite a long time in a maze of fascinating riddles. Today, however, a scientist usually enters the game by personal choice, during training, and aims to work in a particular branch. Conservation science is now an established profession, even if its core is not yet well established and its borders fuzzy.

Other people involved in the conservation trade—conservators, historians, architects, engineers, amateurs, administrators, politicians, and journalists—seldom realize how specialized science for conservation is, or should be. They frequently consider the conservation scientist to be a jack-of-all-trades to whom they turn for the solution of any awkward problem they feel unable to solve (that means, of course, that the problem must be a most untreatable one, because people in conservation usually feel competent to solve anything).

Unfortunately, most scientists, this writer included, are unable to say no—because the postulant is desperate, or there is an exhibition tomorrow, or the object is exposed to a most unfriendly environment and on the point of collapsing. Scientists feel obliged to answer the call, even if the material and the action required are out of their domain of competence. They normally pick up the gauntlet and take a gamble. It may pay off, but sometimes damage is done.

Is Conservation Science Really "Science"?

The fact that conservation scientists occasionally venture out of their field of competence and take risks that can potentially damage precious materials is not the only problem that affects conservation science as a whole. Unfortunately, scientists involved in conservation are gamblers even in their field of competence, because most of the time they offer interpretations and solutions despite insufficient knowledge.

Consider an analyst studying a piece of an ancient monument. In the majority of cases, he lacks the information required for a reliable interpretation of the results. One reason for this is that the object probably has been modified or treated several times, and typically the scientist has incomplete information on its history. Many samples should be taken and compared to separate materials belonging to different phases, but frequently this is not possible, either because of cost constraints or because damage would be inflicted by sampling.

The 1981 testing in the ICCROM laboratory of a grouting mixture, made of hydraulic lime combined with various admixtures, intended for consolidating surfaces of artistic or architectural importance. The grout is injected into a column of sand to test its injectability. Testing also includes an analysis of soluble salts, and the measurement of the flow of the injected grout and of the porosity of the hardened material after injection. Photo: J. Malliet.

Furthermore, several analyses should be made on each sample in order to extract maximum information and to cross-reference the results. This makes the data more reliable, as ancient artistic or architectural materials are normally very complex mixtures that offer no clear-cut result in any single method of analysis. Unfortunately, the application of multiple analyses to the same sample is normally not possible, either because of cost or because of the dimension of the sample.

In such conditions, interpretation will be based on insufficient data and be strongly influenced by the analyst's preconceived notion of what the result should be. Nonscientists often don't realize that in routine analysis, the scientist finds only what he seeks—i.e., what the scientist thinks is in the sample. The chance of unforeseen findings increases when the results of different analyses are compared. This explains, for instance, why calcium oxalates, found by Justus von Liebig on the Parthenon marbles in the 19th century, were not detected again on the surface of ancient stones for a long time, while today—after they were identified again in the 1960s by an intensive use of X ray diffraction—they seem to be almost ubiquitous.

These photographs document the creation of a new conservation technique, illustrating how conservation science operates on the border of different domains—in this case, chemistry, mechanics, and the empirical knowledge of conservators.

Testing the grout on a larger scale in 1982, at the Construction Science Laboratory of the University of Rome. Photo: Giorgio Torraca.

Further testing at the University of Rome in 1982. An experimental wall, partially crushed under a press, is consolidated by grout injection. Photo: Giorgio Torraca.

When the analyst writes a report interpreting data in terms of the history of the object or its state of conservation, he gambles, relying on general experience to extrapolate from the data a presentable hypothesis (which is necessary if the analysis is to be of any use to those paying for it). Another reason for gambling is that the analyst thinks it unlikely that anyone will read the report in detail; to further reduce that possibility, he uses the most abstruse technical jargon in writing it.

The scientist dealing with conservation processes must also gamble. The data at the scientist's disposal for evaluating the cause and rate of deterioration of an object requiring treatment are normally insufficient. The same applies to the evaluation of the future service life of the materials that may be used to consolidate the object under treatment and delay its decay.

Conservation today is a production line—the more so in architectural conservation. It no longer proceeds at the leisurely pace of the still-recent past, when conservation was such a quiet and pleasant profession. Problems must be solved within deadlines that do not allow sufficient time for experimentation and analysis. As a consequence, when a scientist proposes a conservation treatment and guarantees its reliability and durability, he is either consciously bluffing—in the best cases—or suffering from delusions because of lack of experience.

Actually, all branches of science and technology involve some gambling in the creation of hypotheses or models that are not logically deducible from the available experience. However, science proceeds by experimenting in such a way that a hypothesis may be refuted or supported by the data thus obtained. "Progress" in science is achieved by modifying the models, according to the results of experiments, in order to bring them closer to reality; gambling becomes less important as science becomes more mature.

In conservation science, however, the importance of gambling (explaining things by unsupported hypotheses) is much greater than in most domains of science. This may mean that conservation science remains in an early state of development, in which imagination still prevails over hard facts. But there are also reasons to suspect that conservation science may not—or may not yet—be a "science," as defined by contemporary thinkers, even if it employs scientific equipment and scientific language.

The main reason supporting such a statement is that it is difficult to falsify or support a hypothesis (or model) in conservation.

If "scientific" research is directed toward understanding the history of an object, it is difficult to prove or disprove any hypothesis about an object's past on the basis of the "scientific" data alone. Help may come from other disciplines. There may, for instance, be the discovery of that rare document whose interpretation is univocal. But in the typically uncertain domain of history, positive proof is the exception rather than the rule.

If we are dealing with conservation treatment technology instead, any hypothesis that is advanced can lead to the practical consequence of conservation treatment, and the result—for example, the object's decay rate after conservation—might be measured.

In such a case, a model could be supported or disproved. However, it takes a great deal of time—probably decades or even centuries—to confirm that long-term conservation has been achieved. While falsification or support for a hypothesis is possible in principle, it can only occur after a lengthy period. Scientific progress is therefore bound to be very slow.

Between Two Cultures

Conservation scientists might be quite annoyed to be told that their discipline should not be considered "real" science. Still, the idea that its nature is to straddle the frontier between two different cultures has positive elements in it.

In the first place, this middle position helps explain some facts about science applied to conservation that keep disturbing us. These facts would be viewed more leniently if they were considered to be normal consequences of conservation science's borderline status.

An example of such disturbances is the frequent occurrence, in conservation, of inaccurate analyses and unreliable testing of materials.

I remember that many years ago Tom Chase of the Smithsonian Institution led the ICOM committee for metals in an experiment of interlaboratory analysis using a homogeneous sample that he had made by grinding an ancient piece of bronze into a fine powder. The scattering of results he obtained from museum laboratories was far beyond the limit considered acceptable for industrial laboratories. (Nowadays, we protect ourselves against similar discouraging finds by never repeating an analysis.)

If the analytical data are unreliable, their interpretation is bound to be more unreliable by an entire order of magnitude. For example, a few years ago, a well-known mineralogist declared authentic some Modigliani sculptures retrieved from a ditch (where the sculptor presumably threw them). The basis for this conclusion was the finding that the layer of mud in contact with the stone sculptures contained almost no lead (evidence that there were no cars and no gasoline with tetra-ethyl-lead in it at the time the mud layer was formed). Contrary to what normally happens with archaeometric interpretation, the gambler in this case was unlucky. The forgers were there and able to prove that they had made the presumed masterpieces a few days before their discovery.

Using the grouting technique in the field in Pompeii in 1983. Photo: Giorgio Torraca.

Consolidation in 1985 in Assisi of a mural painting, damaged by earthquake. New conservation techniques always carry risks, even when extensive testing precedes fieldwork. In the case here the grouting technique has thus far proven successful. Some commercial products, with composition similar to the ICCROM grout, are now widely used in the conservation of mural paintings, mosaics, and stones. Photo: Giorgio Torraca.

I also think that the testing of materials to be used in conservation is strongly conditioned by the preconceived idea that the scientist has of what the result should be. In fact, most testing reports, aimed to justify the use of a given consolidant or protective material in a conservation process, look very much like the televised claims that a certain detergent washes whiter than another detergent. There is no need, however, to be overly pessimistic. If we concede that conservation science may not be "science," this does not exclude the notion that it may be quite useful (just as detergent technology and advertisement are useful, even if they are not "science").

I would even go beyond such a purely defensive statement to assert that the fact that conservation science is not entirely scientific makes it more interesting, at least to people like me, who think that reality should not be explained only by numbers and formulas. Rather, it should be explained by models that include—besides the numbers (which should be decently correct, if possible)—a lot of material of a different type (words and images) produced from the other culture, the humanistic one.

Since conservation analysts need to consider historic data in carrying out their daily jobs, they should become, little by little, expert also in the techniques used in that domain—archival research, reading of ancient documents—and should acquire a broad view of culture, including social and political history and the history of technology. This would also help the historian; a complete understanding of the meaning of an ancient technical document may be obtained only when it is also read by a technical expert. The professional life of this kind of analyst would be far more attractive than that of the "real" scientific analyst. The conservation analyst will be quite a useful person, even if his or her analyses will probably be less accurate.

Another important consequence of being imperfect scientists is that conservation scientists have a chance to speak and write in such a way that an architect, art historian, or conservator may understand what they have to say. Such an understanding, obviously essential for the success of any enterprise in conservation, is seldom achieved today.

But let's leave the conservation scientist, tinkering on the borderline between two cultures with his high-tech equipment and hyperspecialized language, because on the subject of mutual understanding, something should also be said to the people inhabiting the region beyond. Archaeologists, architects, and art historians, too, should be trained to move into the frontier between culture and science, as conservators already do. These professionals should learn at least enough about science to allow them to look through the scientific trappings that adorn laboratory reports and to reach the useful information that may be there. That sort of activity would be an unorthodox one, as it leads out of the normal paths in the humanities and social sciences. But I believe that such an educational effort is essential for historians, architects, and archaeologists who plan a career in the management of cultural property.

Going back to conservation science for a conclusion, I think that there is no reason to be antagonized if it is found not to be "real" science. To the contrary, if all people in the profession are able to acknowledge this, I am sure that they would be happier and more efficient. The outlook for the artistic and historic property entrusted to their care would be much brighter, and a domain of science- culture would grow, with benefits for both sides of the borderline.

Giorgio Torraca, a former deputy director of ICCROM, is an associate professor in the faculty of engineering at the University of Rome "La Sapienza," and works as a consultant for materials science as applied to conservation.