Abstract
Material deposits disclose their age by exhibiting a condition. Weathering, patina, wear and decay are legible, and a practised eye reads them. Symbolic deposits disclose nothing of the kind, because a distinction carried by a mark does not weather with the mark. The inscribed date is the device that repairs this asymmetry: it gives a made thing a stated position in a sequence rather than an exhibited condition, and it is the only means by which an artefact can declare when rather than merely show how worn. This article traces the emergence of that device through three moments in European painting, from the manufacturing record to the witness statement to the double coordinate of Anno Domini and Anno aetatis suae, and argues that the double coordinate was transplanted from the tomb, where it recorded a completed life, onto the living portrait, where it records a life still running. The transplant converted a terminal sum into an open interval, which is what made annual self-dating thinkable. The article then argues that this pathway to calendar age is structurally invisible to the burial-record evidence on which the standard periodization rests, and closes with the modern inversion, in which the temporal signature has passed from an assertion someone chose to make into a condition no deposit can decline.
I. The Silence of the Deposit
A made thing carries its history in its body. A stone step worn into a hollow, a handle darkened where hands have held it, a beam checked and split, a painted surface craquelured and yellowed: each of these exhibits a condition from which elapsed time can be read. The reading is approximate and requires skill, but it is available, and it is available without anybody having intended to provide it. Material deposits announce their age whether or not they were meant to.
Symbolic deposits do not. A distinction carried by a mark does not weather with the mark. The parchment yellows and the word does not. A classification stabilized three centuries ago arrives in a document exactly as it arrived when it was written, and nothing in its form indicates the interval. This is the source of its enormous usefulness, since a distinction that stayed the same could be carried across absence, delay, distance and the death of everyone involved in making it. It is also the source of a characteristic difficulty. The most portable component of a symbolization carries no temporal signature at all, and the component that does drift, namely what the symbol does when someone enacts it, leaves no external record against which the drift could be measured.
This asymmetry is worth stating plainly, because the rest of the article follows from it.
A material trace exhibits a condition. A temporal signature states a position. Nothing about weathering can tell you which year; nothing about a word can tell you anything at all.
There is a device that repairs the asymmetry, and it is so ordinary that its invention is easy to overlook. It is the date written on the object.
An inscribed date does something no amount of material legibility can do. Wear indicates that a long time has passed. It cannot indicate how long, and it certainly cannot indicate when, since the same wear on the same object would look identical whether the object was made in one century or another. A date states a position in a sequence. It is exact where wear is approximate, it is absolute where wear is relative, and it survives conditions under which the material evidence would be worthless, since a date on a perfectly preserved object still tells you when.
The date is therefore a joint operation of the two poietic mediations. The object persists materially, which is what multimateriality does. The inscription states when the persisting began, which is what multisymbolization does. Neither alone would produce the effect. A dated thing that did not last would tell nobody anything; a date without a bearer would refer to nothing.
This article traces the emergence of that device in European painting, where it can be dated with unusual precision, and argues that its emergence has been misperiodized because the sources ordinarily consulted cannot see the pathway by which it arrived.
II. What a Date Does
Before the historical argument, the operation itself should be specified, because the ordinariness of dating conceals several distinct things it accomplishes.
The first is individuation. A dated object is distinguished from every other object that resembles it, including objects made by the same hand in the same manner. Two panels of similar subject and quality are, until dated, interchangeable in a specific sense: either could occupy the position the other occupies in an account of what happened. A date makes one of them earlier and forecloses the substitution.
The second is ordering. A dated object can be placed in sequence with other dated objects, and the sequence supports inference. Development, influence, revision and priority all become discussable in a way they are not among undated things, where the ordering must be reconstructed from stylistic evidence that is itself contested.
The third is interval measurement. Once two things bear dates on the same calendar, the distance between them is a quantity. This is not available from material evidence, which supports at most a judgement that one thing looks older than another.
The fourth is the one that matters most here, and it has no equivalent in the material register. A date fixes the object's own position relative to whoever encounters it. An undated object is encountered as present: it is here, in this room, now, and however old it is, its age is a property it carries rather than a relation it establishes. A dated object establishes a relation. It says that it began at a point, and it thereby places the person standing in front of it at a measurable distance from that point.
That fourth function is why the inscribed date is not simply an administrative convenience. It installs historical distance into the encounter with an object, and it does so whether or not anyone wants it there. A viewer who reads 1532 on a painted slip of paper has been told, before doing anything else, that what they are looking at was made at a determinate remove from where they are standing, and that the remove can be counted.
An object without a date offers no such relation and can therefore be encountered in other ways. It can be treated as continuous with the present, as belonging to no particular moment, or as a substitute for something older that it reproduces. That last possibility is what the following section takes up, because the refusal of a temporal signature was not an absence of the technique. It was, for a long time and in a particular class of objects, a deliberate choice.
III. The Gold Ground as Refusal
The clearest evidence that dating is a choice rather than a default is that a whole category of objects systematically declined it, and declined it for stated reasons.
Medieval devotional painting placed sacred figures against a gold ground, and the gold ground is a precise pictorial statement rather than a decorative convention. Gold represents uncreated light. It does not cast shadows, does not recede, does not admit atmosphere, and does not belong to any particular time or place. Figures set against it are located outside natural space and outside natural time. This is why the gold ground persisted long after painters had mastered the depiction of three-dimensional space: it was not a technical limitation but a theological instrument. To place a sacred figure in a realistic perspectival space would be to locate that figure in historical time, implying that the figure existed then rather than now.
The temporal logic that the gold ground served was typological. In typological reading, time is not a sequence of unrepeatable events each sealed off from the others by the irreversibility of the past. It is folded, so that an Old Testament event signifies both itself and the New Testament event it prefigures, and the two are bound by a divine intentionality that runs across the sequence rather than along it. Abraham's near-sacrifice of Isaac is not merely similar to the Crucifixion. It is the Crucifixion in prefigural form. This was underwritten by the doctrine of eternity as nunc stans, the standing now, in which all moments of history are simultaneously present to God, who does not remember or anticipate but perceives.
An image serving that logic must not be dated, and the requirement is strict rather than incidental. A date would locate the depicted event at a determinate remove from the viewer, and the entire purpose of the image was to abolish that remove and let the viewer participate in an eternal reality rather than observe a historical one. Anachronism of dress, such as fifteenth-century garments on Mary and Elizabeth, was not ignorance. Collapsing historical distance was the point.
The Konrad Witz workshop panel Der Ratschluss der Erloesung shows this with unusual clarity, since it places the pre-temporal Council of Redemption and the historical Visitation in a single continuous pictorial space with no boundary marked between them. One belongs to eternity and one to the reign of Herod the Great, and the panel does not distinguish them, because under the theology it serves they are the same event seen from two positions.
The relevant point for this article is what such an object cannot have. It cannot bear a date without contradicting itself. A dated devotional panel would be asserting exactly the historical positioning that the gold ground exists to deny.
The gold ground is not the absence of a temporal signature. It is a refusal of one.
That refusal is what makes the appearance of the signature historically legible. Something had to change for a painted object to begin asserting when it was made, and the change was not the acquisition of a technique that had been unavailable.
IV. Three Positions of the Signature
The device arrives in stages, and three objects mark them.
Nicola Pisano's pulpit in the Baptistery of Pisa bears an inscription giving the date of completion, 1260. This is a manufacturing record. It states when the object was made, in the manner of a foundation stone, and it makes no claim whatever about the temporal status of what the object depicts. The sacred figures carved on the pulpit exist in eternity; the pulpit was carved in 1260. The two facts do not interfere with one another, because the date attaches to the fabrication rather than to the depiction. A great many medieval dates are of this kind, and they should not be mistaken for historical consciousness. They are the equivalent of a maker's mark with a year attached.
Jan van Eyck's inscription on the Arnolfini portrait, Johannes de Eyck fuit hic 1434, is a different operation. It is a witness statement. What is being dated is not the manufacture of a panel but an act: someone was present, and was present in a particular year. The date has migrated from the object to an event, and the event is the painter's own presence at a scene. This introduces something the manufacturing record does not have, since it asserts that a specific, unrepeatable occurrence took place and that the painting attests to it.
Hans Holbein the Younger's portrait of Georg Gisze bears the third form. A painted slip of paper carries two inscriptions together: Anno aetatis suae XXXIIII and Anno Domini 1532. Two counts are stated, on two different scales, about the same person at the same moment. One locates the sitter on the calendar of the world. The other locates him within his own elapsed lifetime.
The progression across these three is a progressive intensification of what the date attaches to. First the object, then the act of making, then the body of the person made present. Each step makes the particular more irreplaceable, and by the third step the object has become unable to stand for anything but itself. A panel that says a named man was in the thirty-fourth year of his age in 1532 cannot be a substitute for an older image, cannot be encountered as belonging to no particular moment, and cannot participate in the typological folding that the gold ground protected. The temporal signature has become total.
What the sequence does not explain is where the third form came from. The manufacturing record is an obvious extension of the maker's mark, and the witness statement is an obvious extension of a signature. The double coordinate is neither. It requires that a person's elapsed lifetime be treated as a quantity worth inscribing, and it requires a convention for inscribing it. Both existed already, in a place nobody looking at portraits would think to check.
V. The Formula's Home Is the Tomb
Anno aetatis suae was not invented for portraits. Its ordinary habitat, throughout the medieval and early modern periods, is funerary.
The phrase appears on gravestones and monuments, abbreviated variously, and it appears there in a standard construction paired with the year of the Lord: someone died in a stated year of the Lord, being in a stated year of his age. The two coordinates are given together, as a matter of course, on the memorial of a dead person. The formula answers the question a memorial exists to answer, which is who lies here and how long they lived.
The convention has deep antecedents. Roman funerary inscriptions habitually recorded years lived, with the standard formula stating that the deceased lived so many years, and for children the months and days as well. Soldiers received a second count for years of service. The practice of inscribing an elapsed lifetime on a durable memorial is therefore very old, and it is continuously attested in exactly the material register where such inscriptions survive best.
This has a direct bearing on the Holbein cartellino, and it is the article's central historical observation.
The double coordinate did not have to be invented in the sixteenth century, because it already existed. Anno Domini paired with Anno aetatis suae was a formula in regular use, with an established form and an established occasion. What happened in the portrait was not the invention of a device but the migration of one, from a memorial for a dead man to a likeness of a living one.
The double timestamp was not new in 1532. What was new was putting it on somebody who was still alive.
The significance of the migration is easy to underestimate, because the words are identical in both settings. They are not doing the same thing.
Before turning to the difference, one further feature of the funerary setting deserves notice, because it explains why the formula was available to be borrowed at all.
A memorial is an object whose entire purpose is to survive the person it commemorates and to say something about them to people who did not know them. It is, in the terms of this article, a deposit constructed specifically for temporal signature. Everything about it is designed to state a position: who, when, how long. Of all the classes of made object in medieval and early modern Europe, the memorial is the one in which dating was never controversial, because a memorial that did not state when would have failed at the one thing it was for.
The formula therefore developed in the single setting where the temporal signature met no theological resistance. Devotional images refused dates because dating contradicted their purpose. Memorials required dates for the same reason, from the opposite direction. This is why a fully developed convention for inscribing two coordinates was lying ready, worked out and standardized, at a moment when portraiture had no such convention of its own and needed one.
VI. From Closed Total to Open Interval
On a tomb, a stated age is a completed sum. The person lived that many years and no more, the count has stopped, and the number is final in the strictest available sense. Nothing further will be added to it. That is precisely why it can be inscribed: a memorial is made after the fact, and the fact it commemorates is the closure of a life.
On a portrait, the identical formula states something structurally different. The sitter is in his thirty-fourth year and will, in the ordinary course, be in his thirty-fifth. The number is a position within a process that is still running, and it will be wrong within a year. It is a slice rather than a sum.
The transplant therefore performs a conversion, and the conversion is what makes everything that follows possible.
On a tomb the number is a terminal sum. On a portrait it is an open interval. The transplant converted a completed count into a running one.
Several consequences follow immediately from that conversion.
The first is that a running count implies subsequent counts. A completed sum is stated once and stands. A slice through an ongoing life invites the question of what the next slice would say, and the question has an answer that arrives annually. Whatever the eventual social history of birthday observance, the conceptual precondition is a number that increments, and a terminal sum does not increment.
The second is that a running count requires a known origin. A completed sum can be estimated at the point of death from whatever the household remembers, and the estimate will never be tested. A running count must be recomputable, which means the origin point has to be held somewhere retrievable. The portrait inscription is therefore evidence not only of a convention but of a record.
The third is that a running count makes the number comparable to something the memorial never had to face. A tomb states an age with no body present to check it against. A portrait states an age directly above a painted face. The two are placed in the same object, and a viewer holds both at once.
That last consequence is the subject of section VIII, and it is the one that connects this history to the present.
There is a fourth consequence, less tractable than the others but worth registering. A formula does not arrive from its previous habitat entirely clean. Anno aetatis suae had been, for centuries, what one wrote about a dead person, and putting it on a living one carries a residue of that. The sitter is being inscribed in the manner of a memorial while still alive, which is close to what a portrait is anyway: an object commissioned by someone who knows it will outlast them, made to say who they were to people who will not have met them. Dürer's self-portraits and the Gisze panel are all, in this respect, premature memorials.
That reading should be offered lightly, since it cannot be established from the inscriptions themselves. But it makes sense of something otherwise puzzling, which is why the funerary formula was available for borrowing without anyone apparently finding the borrowing strange. If a portrait was already understood as an object addressed to posterity, then writing on it what one would write on a tomb was less a transgression of categories than a recognition of what the object was for.
VII. The Apparatus Before the Precision
A detail in the formula deserves attention, because it complicates any account that treats the inscription as straightforwardly exact.
Anno aetatis suae XXXIIII does not mean thirty-four years old. It means in the thirty-fourth year of his age, which is to say that thirty-three years have been completed and the thirty-fourth is under way. The construction counts inclusively, in the manner of an ordinal rather than a cardinal, and the person it describes would answer thirty-three to a modern question about their age.
This is not a triviality, and it should not be smoothed over in the interest of a cleaner narrative. It indicates that the apparatus of exactness arrived before the exactness did. The inscription has the form of a precise statement, it occupies the position a precise statement would occupy, and it is placed on the object with all the authority of a Latin formula. What it actually delivers is a number whose interpretation requires knowing a convention, and which is systematically one greater than the number a modern reader would supply.
The same lesson appears elsewhere in the history of calendar age. Reported ages in early modern populations cluster heavily on multiples of five and ten, which is the signature of estimation rather than counting. That clustering coexisted for centuries with the availability of exact formulae, which means the formula and the knowledge were not the same thing and did not arrive together.
A convention for stating an exact age is not the same as knowing one, and the convention arrived first.
This matters for the argument because it blocks a tempting inference. One might suppose that the appearance of Anno aetatis suae on a portrait demonstrates that its sitter possessed and relied upon a precise, recorded, retrievable birthdate of the kind a modern person has. It demonstrates less than that. It demonstrates that a form existed for stating elapsed lifetime as a number, that the form had migrated from mortuary to secular self-presentation, and that a wealthy merchant had reason to use it about himself. What lay behind the number, whether a parish register, a family book, or a household reckoning, is a separate question, and it would have varied.
The correct claim is accordingly modest and still substantial. What the portrait shows is the practice of self-dating by elapsed lifetime, not the infrastructure that would later make such self-dating universal and verifiable.
VIII. Two Counts on One Slip of Paper
Return to what the Gisze cartellino physically is. It is a painted representation of a piece of paper, attached to the pictorial surface, bearing two numbers on two scales, positioned within a portrait of a face.
The face supplies one kind of evidence about the sitter's temporal depth. It is the evidence any human being reads from any other, integrating skin, hair, posture, bearing and a great many cues that nobody can enumerate, and it yields an impression rather than a quantity. The inscription supplies another kind: a count, exact in form, corresponding to nothing visible.
These are different objects and there is no reason for them to agree. One is a reading and the other is a count. The reading cannot be corrected against the count, since it is not an estimate of it, though it can be pressed into service as one. The count cannot be mistaken, since counting is exact, and it also corresponds to no bodily property whatsoever. Nobody has an organ that is thirty-four.
The portrait puts both in one object, and this is what makes it a hinge.
The dated portrait is the first artefact in which a stated chronology and a discernible embodied depth are placed side by side and left for a viewer to hold together.
The comparison the object invites is the one that a later culture would industrialize. It is the comparison that produces the compliment and the insult about looking younger or older than one's years, that sustains an entire economy devoted to altering the signs from which depth is read, and that operates continuously in any environment where a stated number and a photographed body are presented together for assessment. None of that is available to a society lacking either component. The dated portrait is where both first appear on the same surface.
There is a further wrinkle that the modern parallel makes visible. The portrait is a deposit. It records an embodied state at the moment of painting and persists unchanged afterwards, which means that within a few years it presents a body its sitter no longer has, while the inscription continues to state an age he is no longer in. The object was accurate once and is thereafter a record of a moment that has passed, and it makes no distinction between these two conditions. Every dated likeness is on its way to becoming a portrait of somebody who used to look like this.
This is what the temporal signature accomplishes and also what it costs. By stating a position, the object gains the ability to say when. By the same operation it acquires an expiry, since a thing that has declared its moment can be measured against the distance travelled since.
IX. What the Burial Record Cannot See
A periodization question now becomes unavoidable, and it can be stated precisely.
Current research on the rise of calendar age locates the emergence of age reporting in calendar years in southern Germany and northern Italy between roughly 1650 and 1850, using a sample of burial records, and attributes the change to Enlightenment notions of accuracy together with institutional demands such as the requirement that young men prove birth legitimacy to enter guilds. The work traces how calendar age grew into a category by which people were identified.
The Gisze inscription is dated 1532. That is more than a century before the earlier boundary of that window, and the practice it exhibits, stating one's own elapsed lifetime as a number alongside the year of the world, is unmistakably present.
The evidence runs earlier still, and in a form that makes the self-declaratory character unmistakable. Albrecht Dürer's Prado self-portrait of 1498 carries a German inscription recording that he painted it after his own likeness when he was twenty-six years old. His Munich self-portrait of 1500 carries a Latin one naming him, stating that he painted himself thus in his own colours, and giving aetatis anno XXVIII. Two self-portraits, two years apart, in two languages, each stating the painter's elapsed lifetime as a number attached to his own face.
These are worth more to the argument than the Gisze panel, for three reasons. They are self-portraits, so the declaration is not a merchant's commission but a painter's statement about himself, and there is no patron to whom the choice can be attributed. They appear in the vernacular as well as in Latin, which indicates that the practice was not confined to a learned formula being copied. And they precede 1500, which places the practice a full century and a half before the earlier boundary of the standard window.
The conclusion to avoid is that the research is wrong. It is not, and the apparent conflict dissolves once the two bodies of evidence are described accurately.
A burial record contains an age assigned to a person by other people, after death, for the purpose of administrative identification within a parish or civil system. It is compiled by a clerk from what the household reports. It concerns an entire population, including the illiterate and the poor. And it belongs to a system whose purpose is registration.
A portrait inscription contains an age declared by a living person about himself, in an object he commissioned, for the purpose of self-presentation. It concerns one wealthy merchant in a mercantile elite with international connections. And it belongs to no system at all.
These are not the same practice, they have different drivers, they involve different populations, and they leave different traces. A study designed around one of them will not detect the other, and its periodization will accordingly reflect the moment at which its own source type begins to register the phenomenon.
A study that samples burial records will date the arrival of calendar age to the moment calendar age enters burial records.
That is not a criticism of the method, which is appropriate to the question it was designed to answer, namely when age reporting became a general administrative practice for identifying ordinary people. It is a caution about what the resulting date licenses. It licenses a claim about the generalization of calendar age. It does not license a claim about its origin.
The painted evidence supplies what the burial record cannot: a pathway that ran through elite self-representation, in life, more than a century earlier, and that left its traces in objects rather than in registers.
X. Two Pathways, Not One Delayed Arrival
The relation between these two pathways requires care, because the obvious reading, that the portrait practice was an early version of the administrative one which the rest of society eventually caught up with, is probably wrong.
Consider what each was for.
The administrative pathway exists because institutions need to identify and sort people who are not personally known to those doing the sorting. Guild entry, conscription, inheritance, poor relief, taxation and eventually schooling all require a criterion that can be applied uniformly by a stranger. Calendar age is superbly suited to this because it is indifferent to the body, unarguable, and cheap to check once a register exists. Its virtue is precisely its disconnection from anything perceptible.
The self-declaratory pathway exists for the opposite reason. A merchant commissioning a portrait is not being sorted by anybody. He is asserting something about himself, and the number is doing rhetorical work: it fixes him at a moment, in a career, at a stage of a life whose trajectory is part of what the portrait is about. The number is attached to a face, in an object designed to be looked at, and it means nothing without the face.
Two pathways with such different purposes are unlikely to stand in a simple relation of precursor and successor.
Calendar age arrived twice, for different reasons, in different populations, through different media, and the two arrivals should not be collapsed into a single delayed diffusion.
Several predictions follow, and they are testable in principle. The self-declaratory pathway should appear first in populations with both literacy and a reason to memorialize themselves, which points to mercantile, humanist, ecclesiastical and courtly settings. It should appear in objects rather than in documents, which points to portraits, family books, inscriptions and dedications. It should be uneven and idiosyncratic, since nothing compelled it. And it should show no particular correlation with the administrative pathway's chronology, because nothing links them.
Whether the self-declaratory pathway contributed anything to the administrative one is an open question worth asking rather than assuming. A plausible negative answer is available: the two may have run alongside each other with little contact until the administrative one became compulsory and swallowed the other. A plausible positive answer is also available: elite familiarity with exact self-dating may have made the administrative demand intelligible when it arrived, so that what officials asked for was already a thing some people knew how to say about themselves. The evidence needed to decide between these is not the evidence either literature currently collects.
XI. What a Date Forecloses
The temporal signature is ordinarily described as an addition, something an object gains. It should also be described in terms of what it removes, because the removal is what the gold ground was protecting.
An undated image can substitute. It can stand in for an older image it reproduces, it can be treated as a continuation of a sacred original rather than as a new object, and it can occupy a position in a devotional practice without occupying a position in a chronology. This capacity has been described as a principle of substitution, in which the image points away from its own material fabrication toward an atemporal origin, and it is not a deficiency in medieval image-making but a functional property that such images were made to have.
A dated image cannot do this. The date fixes it to a moment of fabrication and thereby forecloses the referral. It becomes, unavoidably, an object made then, by someone, in a year, and its relation to whatever it depicts becomes a relation of representation rather than of substitution.
This is why the arrival of the signature is not simply a gain in information. It is a change in what images are able to be. And it explains why the change was uneven and contested rather than a straightforward improvement that spread as soon as it was available. Devotional painting for religious communities went on refusing dates long after portraiture had embraced them, and it did so because the refusal was doing work that a dated object could not do.
The unevenness could run through a single workshop. Witz's own autograph painting placed a biblical scene on the identifiable shoreline of Lake Geneva, with the Petit Salève and Mont Blanc rendered specifically enough for modern identification, and signed and dated the frame. The same workshop, at nearly the same moment, produced a panel with a gold ground in which eternity and history share a continuous space. Two temporal orders held open within one working orbit, because two kinds of patron wanted two different things from an image.
What this suggests generally is that the temporal signature is a functional trade rather than a technical advance. An object that states its position acquires everything section II described and loses the capacity to be encountered as belonging to no moment. Whether that trade is worth making depends entirely on what the object is for.
XII. Distance as Affordance
A larger claim about historical consciousness lurks here and should be kept within bounds.
Perspectival construction organizes visual space around a single viewpoint at a single moment, and it has been argued that this constitutes a symbolic form, a way of organizing experience that both reflects and constitutes a mode of consciousness. Something similar might be said of the temporal signature: it organizes an object's relation to its viewer around a stated interval, and thereby installs measurable distance where none had been.
The claim should be made as one of affordance rather than causation. Perspective did not cause historical consciousness, and neither did the painted date. What each did was make available a way of organizing experience that had not previously been ready to hand, alongside other developments, textual, administrative and devotional, that supplied historical consciousness with its other supports. An affordance is neither necessary nor sufficient. It lowers a cost.
The cost lowered by the temporal signature is specific and worth naming. Reconstructing when something was made, from its material condition and its style, is difficult, uncertain, contested and available only to specialists. Reading a date is available to anyone who can read. The signature converts a scarce interpretive achievement into a universally available fact, and by doing so it makes chronological placement the default way of encountering objects rather than an option requiring effort.
That conversion is the mechanism by which historical distance becomes ordinary rather than learned. It does not require anyone to have acquired a new faculty. It requires only that objects begin telling people something they previously had to work out.
The corollary is that the demand for dates and the supply of them reinforce each other, and neither has to come first. Objects that state their positions make it normal to expect a position, and the expectation makes it normal for objects to state one. This is a plainer account of how a habit of mind spreads than any that requires a change in perception.
XIII. From Assertion to Condition
The trajectory has a terminus in the present, and it inverts the whole history.
Dürer wrote his age beside his own face. Van Eyck painted his date by hand, and the act of painting it was a choice with a purpose. Holbein painted a slip of paper bearing two counts, and the illusion of an attached document is itself a claim about documentary authority. Every date on every object considered in this article was put there because somebody decided that the object should state its position.
No such decision is now involved. Every photograph carries an exposure timestamp in its metadata, recorded automatically at capture, along with a great deal else. Every file carries a creation date and a modification date. Every message carries a sending time. Every posted item carries a relative interval that updates as it recedes. The temporal signature has become a property of the medium rather than a statement made in it.
The temporal signature has passed from a claim someone made into a condition no deposit can decline.
Three things follow that are worth stating.
The first is that the signature has detached from intention, which means it no longer indicates anything about what the maker wanted the object to say. A painted date asserts. A metadata field records. The interpretive difference is considerable, since an assertion can be examined for motive and a recording cannot.
The second is that the signature has become strippable and forgeable in ways that a painted inscription was not. Metadata can be removed, and its removal is itself informative. Metadata can be altered, and the alteration is ordinarily undetectable from the file alone. A date that nobody chose to add can be taken away by anyone, which reintroduces at the level of the medium exactly the uncertainty the device was invented to remove.
The third concerns what universal signature does to the possibility the gold ground protected. If every deposit states its position automatically, then no deposit can be encountered as belonging to no particular moment. The capacity to substitute, to stand outside chronology, to be present rather than past, requires that the object not declare when it began, and nothing now fails to declare it. Whether anything is lost by this is a question worth asking rather than answering here, but the loss, if there is one, is not a matter of taste. It is the disappearance of a functional property that images once had and were made to have.
XIV. Conclusion
The argument can be gathered.
Material deposits disclose their age by exhibiting a condition, and symbolic deposits disclose nothing, which leaves a gap that the inscribed date fills. The date is a joint operation of the two poietic mediations, giving a persisting object a stated position in a sequence, and it accomplishes something no material evidence can, which is to fix the object's relation to whoever encounters it.
The device was refused before it was adopted, and the refusal was principled. A devotional image serving a typological theology could not state its position without contradicting its purpose, and the gold ground was the pictorial form of that refusal.
Its adoption proceeded through three positions: the manufacturing record, which dates the object; the witness statement, which dates an act; and the double coordinate, which dates both the world and a body. The third of these was not invented for the portrait. It was transplanted from the tomb, where Anno Domini and Anno aetatis suae had long been paired on memorials for the dead, and the transplant converted a terminal sum into a running count. That conversion is the conceptual precondition for annual self-dating, and it is visible in a painted slip of paper in 1532.
The pathway this reveals is invisible to the sources on which the standard periodization of calendar age rests, since a burial record captures an age assigned by others to a dead person for administrative identification, and cannot capture an age declared by a living person about himself in an object he commissioned. The two are different practices with different drivers, and calendar age is better described as having arrived twice than as having arrived once and spread slowly.
What began as an assertion has become a condition. Van Eyck had to decide to write a year on a painting. Nothing now decides, and everything is dated. The device invented to let a made thing say when has become something no made thing can avoid saying, which is a stranger outcome than either the refusal or the assertion, and one that neither the painters who declined the date nor the painters who insisted on it would have recognized.