The Importance of Dating

013

Contacts with history in high school or college have left most of us with something of a distaste for chronology. At least those in the over-thirty generation can hardly have escaped history courses where the instructor concentrated almost exclusively on chronological structure, key events and persons of the period; and the study of history boiled down to memorizing a chronological framework, the dates of kings, and dynastic charts. Does it really matter whether Columbus discovered America in 1392, 1492, or 1592?

From one perspective, precise chronology is not essential to historical appreciation and understanding. To one without a clear perception of the course of events of the 14th, 15th, and 16th centuries, misdating the discovery of America by a century hardly distracts from whatever significance the event has for him.

On the other hand, if a historian were to take such a cavalier attitude toward chronology, he might easily conclude that the discovery of America was the result of the creative forces unleashed by the Reformation or an attempt to test the theory of Copernicus. It is important for the historian to reckon with the fact that Copernicus was nineteen when America was discovered. A historian can do little with persons or events which cannot be fitted into a rather precise chronological framework.

For this reason, if archaeological material is to be of maximum historical value, it must be as precisely dated as possible. For example, an archaeologist discovers a major destruction of a Palestinian town and concludes that the evidence points to a date between 1250 and 1150 B.C. Without further evidence or a more precise dating, it would be impossible to decide whether the destruction was the result of internecine warfare between Canaanite towns, an Egyptian campaign, Israelite tribal conquest, or an attack by a Sea People.

Is archaeology able to provide datings precise enough to be of historical value?

Scholars disagree on an answer to this question. Sober replies vary from “Sometimes” to “Almost never.” One competent scholar concludes that the evidence points to an identification of certain ruins with Saul’s fortress as the best hypothesis. Another insists that the evidence is too meager to justify such a postulation. Such positions tend to become overpolarized when the weaknesses and tentativeness of the best hypothesis are not emphasized or when the “best” hypothesis is ignored as one among several interpretive options. What is a desirable modus operandi? It seems clear that when the “best” hypothesis seems improbable or unconvincing, it might well be dismissed and the material left a conundrum. In other cases it seems desirable to define the best interpretation of the evidence, even if it is necessary to stress the speculative nature of the postulation.

I tend to feel that archaeological material is sometimes susceptible of sufficient chronological precision to be of historical importance. After all, the evidence of ancient history is so limited that very few statements approach the indubitable. To illustrate, Ahab was killed in a battle near Ramoth-gilead about 850 B.C. The probable site of Ramoth-gilead suffered a major destruction about the middle of the ninth century B.C. The identification of the site is not unquestionable. The destruction could have occurred a few years before Ahab’s death when a cow kicked over a lamp or a few years thereafter when an enemy set his neighbor’s grain heap ablaze. Yet, the best hypothesis based on the extant evidence would connect the destruction with the battle in which Ahab was killed.

Where is the line to be drawn between such hypotheses and the objectionable practice of overcorrelation of Biblical and archaeological material? It takes well-balanced judgment to draw that line appropriately, and no two historians would draw it at exactly the same point. It may be observed that advances in archaeological precision make connections more and more viable. It is now frequently possible to date archaeological groups within a quarter- or half-century. Correlations with such material are certainly more convincing than material dated no more closely than within a century or two—as is the case with the vast majority of archaeological finds in Palestine to date.

Imagine that the world’s first excavation was about to take place in Palestine and you were the archaeologist. You proceeded to dig, carefully separating the artifacts from each layer. In post-dig analysis you observed that frequently succeeding layers seemed to contain a very similar repertory of forms, sometimes new forms appeared alongside those of the preceding layer, and occasionally there were complete typal breaks between successive strata. You were led to conclude that the groups separated by major breaks in typology represented important chronological periods and that new forms in basically similar groups represented innovations within the major periods.

This illustrates, in a simplified way, the two disciplines that must be successfully employed if archaeological material is to be closely dated—stratigraphy and typology. If layers are mixed or if a pit is missed in digging, the typological analysis will prove faulty. If the typological analysis does not concentrate on features distinguished by successive layers, it will not be of chronological significance.

Archaeologists operate with relative and absolute dating. The succession of groups of artifacts by layer from the first excavation in Palestine, suggested above, is in fact a relative chronology. It provides evidence of forms which emerge, change, and disappear at specific points in a sequence of layers representing human habitation. In 1901 Sir Flinders Petrie, Palestine’s first scientific excavator, introduced the principle of sequence dating. He attempted to reduce a mass of Egyptian tomb material to fifty successive stages. Material from early Palestinian digs was related to this relative Egyptian sequence by Petrie and others.

How can relative chronology be converted into an absolute B.C. or A.D. date? Suppose that in your first Palestinian excavation you had discovered three Ptolemaic coins of the early third century B.C. on a floor. This would provide a good indication that all groups of artifacts from layers below the floor predated the coins and all subsequent layers postdated them. Nearly every excavation adds a few such links with absolute chronology. When the sequences from all excavations in Palestine are combined, there are enough links with absolute chronology in most periods to provide quite close absolute dates. As evidence from new digs comes to light, the process of refining chronology continues.

From this perspective it might be expected that in its early history Palestinian archaeology was geared to relative chronology, and, as new evidence appeared, gradually more accurate absolute dates were assigned to the successive phases. In fact the actual progress toward precise absolute chronology in Palestinian archaeology took a much more confusing course. Progress followed a circuitous route for many reasons.

Probably the chief block to orderly progress was the lack of much careful, stratigraphic excavation throughout the history of Palestinian archaeology. Despite lack of precise stratigraphic excavation methods, early excavators made progress in understanding the evolution of forms of artifacts. But these results were thrown into confusion by the seemingly reliable results of subsequent digs which were in fact not very reliable. Finally, in the 1930’s more careful digging and publication began introducing the sequential structure which still stands today.

Another difficulty was the fact that from the start Palestinian excavators tied their finds to the epochs and events of Near Eastern history and archaeology. Once such relationships were made, it became difficult to gain general assent for new proposals, even when they were based on much stronger evidence. Palestinian excavators come out of a strong Western tradition of independent research that often makes prima donnas of leading scholars. The failure of archaeologists to consider the material of their colleagues, particularly those of another nationality, before proposing or proclaiming their correlations is by itself an important factor in the lack of steady progress toward chronological precision.

A third deterrent particularly for the third and second millennia B.C., was the lack of consensus on the chronologies of Palestine’s neighbors. Often the links between relative and absolute chronology consisted of imported artifacts from neighboring countries, where they were tied to the reign of a particular pharaoh or assigned a quite precise temporal span. Following Egyptian scholars, who plump for a high, a middle, or a low chronology for the early pharaohs, Palestinian scholars debate most often not the merits of each case, but which chronology best fits the material from Palestine. Such ad hoc arguments hardly contribute to advances in chronological precision.

A fourth obstacle to systematic development of chronological precision was the neglect of detailed, critical pottery studies. Most scholars were content with knowledge of general lines of ceramic development based on oral tradition. These oral traditions were set down in the excavator’s publications of his pottery. Newer pottery publications then cited or catalogued the previous examples of a particular pot along with proposed datings. Often no distinction was made between dates proposed for unstratified examples and for specimens from a context with good links to absolute chronology. This chorus of uncritical datings had the unfortunate result of gaining for the datings an undeserved confidence. If a dozen archaeologists cite a similar date for a certain form, it must be correct. Unfortunately, in many cases the dozen archaeologists were all merely repeating what was originally an uncritical oral tradition. Advances in chronological precision come when an archaeologist with a rigorously critical approach publishes a detailed study of the stratigraphic changes in ceramic forms and the evidence for their absolute dates.

By far the most common finds on Palestinian tells are potsherds. Broken pieces of jars, jugs, bowls, and lamps litter virtually every layer of the mound from about 5000 B.C., when pottery first appeared in Palestine, down to modern times. In fact, many of the layers in the comparatively poor tells of Palestine contain nothing but sherds. If changes in pottery take place at a fairly rapid pace and can be identified, pottery can serve as the archaeologist’s best chronological tool.

Pottery has many variable characteristics. Preparation of the clay includes levigating it to a coarser or smoother consistency and adding inclusions. The inclusions vary in fineness and consist of straw or grit of various kinds to stone or sand. The clay is formed into vessels by hand, on a slow or fast wheel, or in a mold, or a combination of these. The vessel may be shaved, decorated with incision or puncturing, surfaced with slip, paint, or glaze, and burnished or polished. It may be fired high or low in a kiln with or without reduction or stacking. Further decorative wash, burnishing, or incision may be executed after firing.

Nearly all of these features are of at least potential chronological significance. Pottery from one era of a town’s history may have used different clay beds than their predecessors, used different inclusions, preferred incision to painted decoration, fired their kilns higher, and the like. Some features offer rather precise chronological indications. For instance, pattern burnishing was in vogue only for a few brief periods of Palestine’s history, and overfiring of pots was common in the later Late Bronze age and the Early Hellenistic period. Spectographic analysis and careful attention to inclusions would provide helpful results for chronology, but the analysis is expensive and time-consuming and will be of little significance until such analysis is consistently reported by a number of excavations.

While these characteristics offer chronological data, only rarely do they reach the level of chronological precision required if archaeological material is to be of historical value. The same clay beds, potters’ wheels, and painted traditions often lasted for centuries. There is only one feature of ceramic vessels that undergoes a continuous process of change that can be observed at intervals of a quarter- or half-century. That feature is shape.

The most common categories of ancient ceramic vessels are jars, jugs, juglets, craters, bowls, cups, platters, cooking pots, and lamps. In the Early Bronze age, for example, there may be several jar types but perhaps only one jug type. Some jar types continue through the Early Bronze age; others may be introduced at an early stage and disappear at a late stage within the Early Bronze age. These are chronological facts of significance, but what is of importance for precise chronology is the change that takes place in the shape of a particular jar type during the course of the Early Bronze age. Bases may become more pointed, rounded, flattened, or elongated. Handles may shift their point of attachment, change their manner of attachment, or develop from round to oval to flat. Most often rims are the best chronological indicators. Possible rim shapes are almost infinite, and a particular rim shape tends to develop persistently and rapidly. Such changes in shape are perhaps the closest counterparts to the modern phenomenon of annual model changes for cars and appliances.

To leave the impression that all pots change shape regularly and rapidly would be incorrect. Some shapes tend to change much more rapidly than others. Jar rims tend to be much more precise chronological indicators than lamps. Some forms seem to persist for centuries with little or no change, while other types disappear after flourishing for only a few years or decades. A few simple types tend to persist or recur century after century. Some exotic shapes have a very short life and can be immediately identified and closely dated even by a novice.

For those schooled in an “evolutionary mentality,” it should be noted that the formal trend of Palestinian—and indeed all ancient pottery—is commonly downward. Finely made examples of a new and pleasing shape appear. Soon a decline in quality becomes apparent and the shape begins to sag or bulge. Since it is virtually impossible to accurately publish subtle changes in quality even by verbal description, a refined understanding of the dating of pots requires extensive experience in going over basket after basket of pottery as it comes from layer after layer in the field.

The knowledge of pottery chronology grew as leading excavators such as Sir Flinders Petrie, Père L.-H. Vincent, and Clarence Fisher shared their field observations. The digging of W. F. Albright at Tell Beit Mirsim in southern Judea resulted in a critical examination of this oral tradition. In the final reports of that excavation, published between 1932 and 1943, Albright was able to provide a fairly complete picture of the changes in pottery shapes from the late third millennium down to the early sixth century B.C.¹ Pottery chronology for the periods preceding Tell Beit Mirsim was systematized in a dissertation under Albright’s direction by G. Ernest Wright in 1937.² Systematization of our knowledge of pottery chronology for later periods has been slow to appear. The writer’s dissertation, published in 1961, attempted to systematize the ceramic chronology for the period from 200 B.C. to 70 A.D.³ Well-dated material for the late fourth and third centuries B.C. has been recently published by Fawzi Zayadine and Nancy Lapp,⁴ and the first published attempt to define the formal changes of the Persian period is found in my contribution to the Kurt Galling Festschrift, issued in 1970.⁵ Knowledge of pottery development in the periods after 70 A.D. is still confined largely to a quite imprecise oral tradition.⁶

The constant stream of newly published material has made possible considerable refinement of the chronologies of Wright and Albright. Among the most notable of these are Basil Hennessy’s study of the Early Bronze age, Kathleen Kenyon’s analysis of the Bronze age tombs at Jericho, and the publication of the Iron age pottery of Samaria by Kenyon and of Tell Deir ’Alla by H. K. Franken. In our present state of refinement, and with few exceptions, we are able to date larger ceramic groups from the late fourth millennium B.C. through the first century A.D. within a century. In many instances in the last two millennia B.C. it is possible to date groups within fifty years, and on occasion within a quarter-century.

It should be apparent that an expert in ceramic chronology is indispensible for any dig dealing with post-5000 B.C. material in Palestine. In most cases pottery provides the crucial and exclusive evidence for dating the dig’s discoveries. If this evidence is not controlled, the finds of a dig are of little more significance than those purchased in an antique shop. In fact, if an excavator does not have a thorough knowledge of ceramic chronology, he would be favoring historians if he were to stop digging and purchase his artifacts from an antiquities dealer.

Before the introduction of pottery about 5000 B.C., the most ubiquitous stratified finds are flints. While they provide the best evidence available, they do not provide as precise chronological indications as pottery. Once introduced, a flint type tends to persist longer than a pottery type, but without any developments comparable to those of pot rims, handles, and bases. This means that our chronological knowledge is much less precise, and, in fact, the farther into the past we penetrate, the longer are the periods we are able to identify.

The resulting picture is usually interpreted as an indication that the farther back we go, the slower was the pace of change and development. This interpretation should be accepted with some skepticism, for the picture is actually the result of a lack of materials for delimiting shorter periods of time. Time and again new evidence from pre-history has astounded even the specialists with its sophistication. The third millennium B.C. is still commonly called the horizon of Palestine’s earliest towns, but the Jericho excavations have brought to light a massively fortified town of the seventh millennium B.C. The sophistication of its houses is sufficient to emphasize that periods of creative development were heavily interspersed with eras of stagnation. Some earlier ages may have changed as rapidly as those where pottery makes the changes perceptible.

Technically speaking recorded history begins about 3200 B.C., the date of the Sumerian documents bearing man’s earliest known writing. By the early third millennium B.C. the archaeological materials of Palestine can be linked with the absolute dates derived from the early documents of her neighbors. Until recently, prehistoric material was largely confined to sequence dating. Archaeologists may have rather arbitrarily estimated how long certain changes took or relied on the extremely broad datings of palaeontologists, among whom disagreements over periods of five million years are not uncommon.

Much greater precision has been achieved in prehistoric chronology through Carbon 14 dating. Carbon 14 is a radioactive isotope of carbon with a half life adjusted in 1962 to 5730 ± 40 years. Most available dates were figured from previously accepted half life of 5568 ± 30 years. Others have operated with a half life of 5800 years.

This range of over 200 years in half life determination is already an indication of the relative precision of Carbon 14 dates. To be added to this is a ± factor, usually expressed as a 1 value. With a 1 value the chances are one in three that the actual date lies beyond the proposed date range. It would seem preferable to work with a 2 tolerance in which chances are 21 in 22 of the date falling within the proposed range. This would put dates about 3000 B.C. in a ± 250 year range, 6000 B.C. dates in a ± 500 year range, and earlier dates with a correspondingly larger tolerance factor. Carbon 14 dating is still in process of refinement. Its dates will perhaps require additional slight adjustments, but it has already introduced considerable clarity into the chronology of the later prehistoric periods. Its chief limitations are the necessity of a relatively large sample of carbonized material for destruction and the cost of $100 to $250 per test.

It is a common misconception that Carbon 14 has also made important contributions to chronological precision in the historical period. From the third millennium B.C. on Carbon 14 dates can do no more than corroborate in a very general way the datings derived from the study of pottery and other artifacts. If pottery from a certain stratum points to a date in the first half of the sixteenth century B.C., it is reassuring to have a 2 Carbon 14 date of 1720 ± 175 B.C. On the other hand, if the Carbon 14 date came out 1257 ± 160 B.C., it should more than likely be dismissed as a contaminated sample. Carbon 14 dating is not precise enough to contribute to chronological precision in the historical period.

Compared with pots most other kinds of artifacts are relatively poor chronological indicators. They do not occur frequently enough in stratum after stratum to provide the constant chronological indication needed by the field archaeologist. Their sparsity has also meant a lack of sufficient stratified evidence to define rapid typological changes that might exist in certain kinds of artifacts.

Stone vessels and implements are notoriously conservative. Millstones and grinding bowls persist for centuries and millennia with no apparent change of form. Occasional groups, such as the Hyksos alabasters, can be associated with a particular period, but even in such cases I know of no uninscribed stone vessels that can be dated to any single century with confidence. Fashions in metal tools and weapons changed more rapidly, but no particular specimen can be dated within a century. Blown glass, first appearing in the Early Roman period, is limited by its technique to a fairly small typological repertory and is of limited chronological value.

Finds of clothing fasteners and jewelry of all kinds are still less common, so that characteristics only of broader periods may be identified. Jewelry confronts us with other chronological complications inherent in individual taste, international style, and especially the heirloom factor. Jewelry may be preserved in royal and aristocratic families for centuries. Figurines, including the relatively common fertility goddesses, have similar chronological limitations, enforced by the conservative tendencies of superstition.

Scarabs can on occasion serve as links with absolute chronology when they bear the symbols of one of the pharaohs. Unfortunately carving of scarabs bearing a pharaoh’s name often continued long after his death. As a result, the manner in which the carver portrayed the beetle is often of more chronological importance than the symbols carved on its belly. Recent publications of groups of hundreds of scarabs, such as those from the Jericho tombs, are disappointing from a chronological perspective.

Seals, too, on occasion serve as links with absolute chronology when they bear the name of a known ruler or official. Even when they bear unknown names, their script can often be quite closely dated. Seals belong to artistic styles which can at times be dated within a century. If seals were as common as potsherds on Palestinian mounds, our chronological framework would be much more precise. Unfortunately they are rather rare finds in the poor mounds of Palestine.

Coins first appear in Palestine about 500 B.C. and become common after the fourth century B.C. Since most coins can be attributed to a particular ruler and often indicate a specific year of his reign, they are of considerable importance as chronological indicators and as links with absolute chronology. Perhaps the main reason for the comparative neglect of pottery chronology after the sixth century B.C. is the assumption that coins can replace potsherds as chronological indicators in the later periods. This premise is wrong for a number of reasons. While coins are rather common finds, they are far less ubiquitous than potsherds. Many later layers of Palestinian mounds contain no coins. Since coins are small, they often present stratigraphic problems. Has this coin slipped through a crack into a lower layer? Was this coin on the floor or in the fill immediately above? Even hoards of coins may present stratigraphic difficulties for they are frequently hidden in cracks and crevices. Most important is the heirloom factor. Coins remained in circulation in antiquity longer than they do at present with the popularity of coin collecting. When a single coin is found on a floor, it is impossible to tell for sure whether it was minted a few years, several decades, or even a century before it was lost.

Where ceramic chronology has not been precisely worked out, numismatic evidence often provides the best chronological data. But where groups of pots can be dated within a quarter- or half-century, coin evidence often provides no more than general corroboration of the ceramic dating. The endurance of common crockery for a half century or more would be quite exceptional, but quite common for coins. It is strange but true that uninscribed sherds can often serve as better chronological indicators than dated coins.

As already noted, shorter inscriptions are often found on scarabs, seals, and coins. Occasionally a pot will be incised with the name of its owner or potter. Monumental inscriptions are found on standing stones or stelae, on statue bases, and on the walls and floors of buildings. All such written material may be classed as incised because its letters are ordinarily carved. Related are the clay tablets inscribed in the cuneiform characters of the Mesopotamians by a wedge-headed stylus. Documents written in ink are less often preserved, but such recent discoveries as the Dead Sea Scrolls, the Samaria Papyri, and the Arad ostraca are spectacular examples of the preservation of such documents in Palestine. Most of the Dead Sea Scrolls are parchments, written on specially treated animal skins prepared in long rolls. Papyrus sheets are prepared by weaving strips of the papyrus reed. An ostracon is a potsherd which served as a piece of scratch paper, for short messages, receipts, bills of lading, and the like.

Written finds are even more infrequent in Palestine than in neighboring lands, and documents bearing a date are rare among written finds. Not a single specific date is mentioned in the Dead Sea Scrolls, but each of the Samaria papyri bore the day, month, and year of writing. The writers, of course, were not aware of how many years before Christ they lived, but the years of the reign of a particular Persian king are easily converted into B.C. dates.

Even when documents bear no dates, they are usually of considerable chronological help. This is because handwriting changes about as rapidly as the shapes of pots, and can be closely dated. Even though the Dead Sea Scrolls bear no dates, the evolution of the script of the first centuries B.C. and A.D. makes possible the dating of individual documents to a quarter- or half-century. It should be emphasized that just as the inscriptions on contemporary buildings differ from our handwriting, so it is necessary to distinguish between the development of ancient incised and cursive scripts. Cursive writing tends to change much more rapidly.

For the last three millennia B.C. the most common finds providing links with absolute chronology are imports, mostly pottery. Often the pots imported into Palestine from Egypt, Cyprus, Greece, Phoenicia, Syria, Anatolia, Mesopotamia, and even farther abroad can be assigned quite precise absolute dates in their countries of manufacture. Discounting a short time lag for shipment, these dates may be accepted as manufacturing dates for the examples found in Palestine.

It should be emphasized that the date indicated by the imported pot is its date of manufacture, and the date indicated by a script is the date the document was written or the artifact was inscribed. This does not mean that these dates can automatically be applied to the strata in which they are found. If they indicate a date in agreement with that provided by the local potsherds with which they are found, they corroborate that date. They may well date earlier than the ceramic horizon with which they were found. Dead Sea Scrolls from a single cave may be dated from the late third century B.C. to the middle of the first century A.D. Imported pots may have been used rarely and been preserved much longer than local kitchen ware, just as our “good china” is often much older than our “everyday dishes”. Written documents and imports, like coins, are often subject to the heirloom factor. On the other hand, if the date provided by the import or written material postdates the ceramic horizon of the layer, something is wrong with the dating of the pot in its country of origin, the dating of the script, the local ceramic chronology, or, most commonly, the stratigraphic excavation or analysis.

A very rare link with absolute chronology is provided when constructions specifically described in historical texts can be located. The most dramatic example of this for Palestine is a discovery of Yigael Yadin. After excavating a gateway from the time of Solomon at Hazor, he noted gateways with virtually identical plan and dimension at Megiddo and Gezer (see “Yadin’s Popular Book on Hazor Now Available,” BAR 01:03). That all three were in fact planned by Solomon’s architect is confirmed by the Biblical account, “This is an account of the forced labor King Solomon levied for the building of … the wall of Jerusalem, Hazor, Megiddo, and Gezer” (1 Kings 9:15). This means that the pottery contemporary with the construction of these defenses may be assigned with considerable confidence to the reign of Solomon. Even such discoveries contribute less to the refinement of ceramic chronology than might be expected. Unfortunately, none of the three gateways have been excavated according to the best stratigraphic methods.

Contacts with history in high school or college have left most of us with something of a distaste for chronology. At least those in the over-thirty generation can hardly have escaped history courses where the instructor concentrated almost exclusively on chronological structure, key events and persons of the period; and the study of history boiled down to memorizing a chronological framework, the dates of kings, and dynastic charts. Does it really matter whether Columbus discovered America in 1392, 1492, or 1592? From one perspective, precise chronology is not essential to historical appreciation and understanding. To one without a clear […]