Jan F.

Simek Chronological changes in stone tool

Department of Anthropology, assemblages from Krapina (Croatia)
University of Tennessee, Knoxville,
Tennessee 37996, U.S.A.
This study presents the results of the first recent analysis of stone tool
assemblages from Krapina (Croatia). All assemblages are Pleistocene in age
Fred H. Smith and many are associated with human remains, the Krapina Neandertals.
Department of Anthropology, The assemblages are described typologically and technologically, and subtle
Northern Illinois University, DeKalb, chronological changes in raw material selection and technology of tool blank
Illinois 60115, U.S.A. production are observed. These changes involve increasingly sophisticated and
selective use of lithic materials. Changing artefact assemblages are considered
Received 31 May 1996 in light of variability in the hominids from Krapina, and are interpreted as
Revision received 17 December reflecting behavioral change among Neandertals rather than between
1996 and accepted 18 December Neandertal and modern human populations.
1996 ? 1997 Academic Press Limited

Keywords: Krapina, Mousterian,

Middle Paleolithic, Neandertals,
Pleistocene. Journal of Human Evolution (1997) 32, 561–575

Introduction
The site of Krapina in northern Croatia was excavated between 1899 and 1905 by the
Croatian paleontologist Dragutin (Karl) Gorjanović-Kramberger. It yielded a large series of
lithic, faunal, and human fossil remains, which were originally described and analyzed by
Gorjanović in two monographs (Gorjanović-Kramberger, 1906, 1913) and a series of shorter
publications (see lists in Smith, 1976 and Radovčić, 1988). The human fossils have been the
subject of several subsequent assessments (e.g. Hrdlička, 1930; Smith, 1976, 1982; Wolpoff,
1979; Radovčić et al., 1988); but the stone tool and non-human bone assemblages, so
painstakingly excavated by Gorjanović, have received only cursory attention subsequent to his
1913 monograph on the ‘‘life and culture’’ of the Krapina people.
The fauna from Krapina was re-examined by Malez (1970a, b, 1978) and others (e.g.,
Guenther, 1959), but the focus of these studies was primarily chronological. The lithic
assemblages were also reviewed by Malez (1970c, 1978); but both Malez and Gorjanović dealt
only with the retouched tools, even though Gorjanović collected all lithic items regardless of
whether or not they were modified. In short, the Krapina archaeological samples never have
been afforded the attention they warrant from students of European Paleolithic prehistory.
Recently, the Krapina stone tools were the subject of a reanalysis (Simek, 1991). In that
study, the lithic assemblage from Krapina was treated as a whole, as if no stratigraphic
subdivisions existed. Gorjanović had, of course, defined a series of stratigraphic units at the site
during his excavations (Gorjanović-Kramberger, 1899, 1901), but he recorded provenience
only for the retouched tools and not for the nearly 800 pieces of debitage he recovered. The
goal of the 1991 study was to address relatively simple problems relating to the Krapina lithics,
such as basic typology, lithic technology and raw material use. For such purposes, analyzing
the lithic assemblage at Krapina as a single unit was both justified and necessary. Admittedly,
any chronological variation in the assemblage was obscured by this strategy.
Simek (1991) did little more than present a first typological description of the tools and
preliminary examinations of stone reduction and raw materials. It was possible to reconstruct
a chaîne opératoire that was essentially identical to that proposed by Baumler (1988) for Middle
Paleolithic materials from Zobiste. This technology involved systematically opening river
cobbles in a series of ‘‘cobble wedges,’’ resulting in amorphous flakes with remnant cortex and

0047–2484/97/060561+15 $25.00/0/hu960129 ? 1997 Academic Press Limited

562 . .   . . 

numerous naturally-backed knives. Krapina technology was based on using raw materials
obtained from the nearby Krapinica River, where small cobbles of volcanic tuff, flint, and
silicified limestone were all that was available. A few large artefacts were produced from exotic
materials for which no local source could be defined and of which no waste was present in the
assemblage. This indicates curation of some materials as the Krapina people moved over the
landscape. Still, river cobble exploitation was by far the primary factor determining the lithic
technology used by Middle Paleolithic Krapina occupants.
Chronological change is a major interest at Krapina, because anatomical differences have
been suggested for hominid fossils from different strata. In this paper, we examine the Krapina
stone tool assemblages in a more detailed chronological framework. Our primary goal is to
detect changes that might be related to evolutionary trends in human ability or behavior at
Krapina. First, we look at the stratigraphic divisions defined by Gorjanović and how they
might be used to examine change over time. Second, we look at groupings of the Gorjanović
strata suggested by the vertical distributions of objects. Because of the nature of artefact
provenience, chronological units are compared in rather simple terms focusing on retouched
tools. Typology and raw material technology will be the principle dimensions to be analyzed.
Despite limitations imposed by the data, however, looking at the Krapina lithics over time
enhances our view of Middle Paleolithic site use at this important locality. Finally, we discuss
the chronologically-based patterning of the stone tools in relation to the human fossil remains
from different strata.

The site
The prehistoric site of Krapina is a collapsed cave feature in a sandstone bluff overlooking the
Krapinica River in the Hrvatsko Zagorje region of modern Croatia, 40 km NNW of Zagreb
(Gorjanović-Kramberger, 1899). During the initial year of excavations at Krapina in 1899,
Gorjanović identified nine major stratigraphic levels in the 11 m of Pleistocene deposits
accumulated in the former rock shelter (Gorjanović-Kramberger, 1899, 1900). Many of these,
particularly the ‘‘cultural layers’’ (levels 2–9), contained sublevels and lenses, suggesting a
complex stratigraphic sequence (Figure 1). Despite this apparent complexity, Gorjanović
believed that the Krapina deposits accumulated over a relatively short temporal span, an
interpretation derived primarily from his observation that the sandstone bluff containing the
cave weathered rapidly to produce the sediments filling the cave.
Based on the fauna, Gorjanović argued that the deposits formed during a warm period.
Initially, he suggested a ‘‘Günz-Mindel’’ or ‘‘Mindel-Riss’’ age (Gorjanović-Kramberger,
1906), but his final view placed the deposits in the ‘‘Riss-Würm’’ interglacial (Gorjanović-
Kramberger, 1913). Subsequent considerations of the fauna and stratigraphy (Malez, 1970a, b,
1978) argued for a longer depositional history at Krapina, extending from the end of the
Riss-Würm well into the Würm glacial complex. However, recent electron spin resonance
(ESR) dating of levels 1, 5–6, and 7–9 indicate that the entire Krapina sequence was deposited
at around 130 ka at the end of Isotope Stage 6 or the beginning of 5e, corresponding to the
late Riss glacial or Riss-Würm interglacial (Rink et al., 1995). Gorjanović was remarkably
accurate in his interpretation of the deposits.
Archaeological materials were scattered throughout the eight cultural levels at Krapina,
although there are distinct vertical concentrations. One of these is found in level 3 (also noted
by Gorjanović in 1901) and another in level 8. Most of the hominid fossils were recovered from
levels 3 and 4, designated the ‘‘Hominid Zone’’ (Smith, 1976). The juvenile A cranium
      563

Figure 1. Gorjanović’s stratigraphic profile of Krapina.

(Krapina 1) and a few other specimens (Radovčić et al., 1988) were recovered from level 8,
near the top of the sequence.

The hominid sample

The human skeletal remains from Krapina comprise some 900 specimens, ranging in
completeness from partial crania to fragments of individual bones (cf. Radovčić et al., 1988). In
his excavation notes and early publications, Gorjanović (1899, 1900, 1901) refers to the
human remains as either Homo sapiens or simply as Pleistocene (‘‘diluvial’’) humans. From the
outset, however, he clearly recognized their archaic nature and similarity to known Neandertal
specimens (Gorjanović-Kramberger, 1899). By 1902, Gorjanović formally associated the
Krapina remains with the ‘‘. . . Formkreise des Homo neanderthalensis im Sinne Schwalbes’’
(Gorjanović-Kramberger, 1902: p. 215). After attending a 1903 scientific meeting in Kassel,
Germany, he began to refer to the Krapina remains as Homo primigenius (Gorjanović-
Kramberger, 1904), the taxon generally preferred for Neandertals in pre-World War I
German-speaking Europe (cf. Schwalbe, 1904).
More recent studies of the Krapina hominids have re-affirmed the Neandertal affinities of
the specimens from the Hominid Zone (levels 3 and 4) and those from levels 5, 6 and 7 (Smith,
1976; Trinkaus, 1978; Wolpoff, 1979). Furthermore, in the only such study to date, systematic
differences could not be found between the jaws and teeth from the Hominid Zone, and the
564 . .   . . 

more recent specimens from levels 5 through 7 (Smith, 1976). However, determining the
affinities of the juvenile A cranium (Krapina 1) and other specimens from level 8 has proven
more difficult.
Gorjanović never waivered from his assessment that only Neandertals were represented in
the Krapina skeletal sample, including Krapina 1. During his lifetime, he countered strongly
the argument by Klaatsch & Hauser (e.g., 1910) that specimens of more modern humans could
be identified in the sample (see discussions in Radovčić, 1988; Smith, 1997). In 1958, B. Skerlj
was the first to suggest specifically that Krapina 1 was not a Neandertal, but a more modern
human. This claim was subsequently challenged on the basis of relatively simplistic compari-
sons, which demonstrated that Krapina 1 did not fall out of the range of variation for known
Neandertal subadults of comparable developmental age (Smith, 1976).
Beginning in the late 1970s, more pieces of Krapina 1 were identified and added to the
specimen. By 1988, N. Minugh-Purvis suggested, following an extensive comparative study,
that Krapina 1 ‘‘. . . is considerably more modern than the rest of the Krapina skulls’’ (1988:
p. 338). Compared with appropriate Neandertal subadults, Krapina 1 exhibits a more vaulted
frontal squama, relatively weakly-developed brow ridge, and a more abrupt transition between
the vertical and horizontal components of the frontal squama (Wolpoff, 1995). In all of these
features, Krapina 1 approaches the early modern human condition to a greater degree than is
common for other Neandertal subadults. On the other hand, Minugh-Purvis (1988) also noted
that Krapina 1 exhibits as many Neandertal as early modern characteristics.
Since 1988, more pieces of Krapina 1 have been added to the reconstruction. Minugh-
Purvis and colleagues have focused specifically on determining the affinities of Krapina 1 with
an even more extensive examination of metric and qualitative data from this new reconstruc-
tion, comparing it with data obtained from 30 other late Pleistocene immature neurocrania
and a collection of more recent ones. Results of this study (Minugh-Purvis, 1997; Minugh-
Purvis et al., 1996) indicate that while Krapina 1 is often at the extreme of the Neandertal
range of variation, both its metric and qualitative traits are consistent with its inclusion among
the Neandertals.
It is possible that the morphology exhibited by Krapina 1 actually reflects the beginning of
a morphological shift toward the early modern human condition. The interpretation that
Krapina 1 represents such a transition was first posited by Wolpoff (1980) and supported by
Minugh-Purvis (1988). Alternatively, Krapina 1 may show the effects of some gene flow into
this region from the Near East, where more modern humans are possibly established by 120 ka
(Bar-Yosef & Pilbeam, 1993), before the onset of the last glaciation. However, as Minugh-
Purvis (1988) notes, we do not have an extensive knowledge of ontogenetic variation in
subadult Neandertal or early modern humans. Given this fact, along with the recent analyses
by Minugh-Purvis and colleagues, it may be that the seemingly ‘‘progressive’’ features of
Krapina 1 are simply an aspect of normal Neandertal ontogenetic variability and have no
phylogenetic significance.
In addition, there are two other noteworthy specimens from the upper part of the
stratigraphic series (levels 6–9, or what we will define as the ‘‘Upper Unit’’ below). A scapula
from level 6, Krapina 128, exhibits a ventral axillary groove. This condition is rare in
Neandertals, being found in only two of 17 specimens, but it is common in more recent
Europeans and other modern humans (Smith, 1976; Trinkaus, 1977; Frayer, 1992). Also,
Caspari (1991) argues that a presumably Neandertal occipital fragment from level 8, Krapina
11, lacks a suprainiac fossa. Frayer (1992) notes that this is the only Neandertal occipital
(n=24) to lack this feature. While the morphological patterns of Krapina 1, 11, and 128 are
      565

50

40

30
n

20

10

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 2. Frequencies of retouched tools by stratum. Peaks in frequency (stratigraphic concentrations of
artefacts) are present in levels 3 and 8.

certainly intriguing, in our opinion they are not sufficient to warrant a compelling argument
that the hominids of the upper levels are not Neandertals. Furthermore, it is noteworthy that
other specimens from levels 6 and 7 are morphologically commensurate with a Neandertal
classification (Smith, 1976).

The assemblages
The lithic assemblages from Krapina comprise 1191 artefacts. These are presently dispersed in
three repositories, two in Zagreb and one in Varazdin. Of these artefacts, 273 fit into Bordes’
Middle Paleolithic tool typology (Bordes, 1954, 1961). Eleven flakes show marginal edge
damage, possibly due to utilization. Thirty-eight retouched pieces do not fit in Bordes’ type list,
and 784 unmodified pieces of debitage were identified. Gorjanović marked in pencil all
artefacts that he recognized as retouched tools with their stratigraphic level and a catalog
reference number. These artefacts still bear field numbers in Gorjanović’s hand. A number of
retouched tools not recognized as such by Gorjanović were identified during the studies
reported here. None of the unretouched debitage or recently-identified tools have stratigraphic
information. Thus less than 300 artefacts (or about 24% of the total) can be subdivided
into stratum assemblages. Artefacts were found in all of Gorjanović’s ten geologic levels at
the site.
No single level has anywhere near the 100 tool minimum deemed necessary by Bordes for
typological analysis, but the distribution of tools among levels is informative from a
stratigraphic perspective (Figure 2). Two clear modes are evident in artefact frequencies by
level, one at level 3 and another at level 8. Hominid fossils were also concentrated in level 3
and to a much lesser extent, level 8. Given the age of excavations at Krapina, and what we
understand today about processes of site formation, there may have been fewer archaeological
layers than geological beds at the site. Two levels would seem the minimum number, separated
at geologic stratum 6. A maximum cannot be estimated because we have no provenience
finer than geologic stratum. However, given the rapid rate of sedimentation inferred by
Gorjanović and supported by recent dates, fewer than ten archaeological layers seems
probable.
566 . .   . . 

100
90
80
70
Percent 60
50
40
30
20
10
0
10 20 30 40 50 60
Bordian type
Figure 3. Cumulative frequency graph for Krapina stone tools classified according to the 62 Middle
Paleolithic tool types defined by Bordes (1961). The form of the distribution is typical of the Charentian
Mousterian (i.e., rich in various sidescraper forms).

Thus, in the chronological comparisons below, we use both Gorjanović’s ten strata (the
minimal units available today) and two groupings of levels based on observed distributions.
The two larger groupings, the Lower Unit containing artefacts from levels K through 5 and the
Upper Unit comprising materials from levels 6 through 9, may more closely approximate the
actual assemblages deposited by Paleolithic site occupants. Whether there were few or many
archaeological layers cannot be reconstructed with certainty, and all studies of Krapina
archaeology (including this one) must be viewed with caution with respect to the stratigraphic
units being considered.

Stone tool typology

According to Bordes’ Middle Paleolithic typology, the Krapina assemblage is Mousterian with
a mix of various major tool classes (Figure 3). However, sidescrapers are abundant overall, and
this must be considered as a ‘‘scraper-rich’’ or generic Charentian Mousterian as defined by
Bordes, and more recently, by Geneste (1985). Sidescrapers comprise more than half of the
tools. Notched and denticulated pieces are common, as are naturally-backed knives; Upper
Paleolithic tools are rare. Levallois blanks are present but not common, so it would be
inaccurate to call this a Ferrassie Mousterian. Almost no true Quina retouch was recorded,
although some retouch is rather invasive. Blanks are not systematically thick as is normally
found in Quina assemblages according to Turq (1992), and this assemblage should not be
characterized as Quina Mousterian.

Raw materials
Lithic raw materials were classified based on petrographic studies of the rocks performed by
Zupanič in 1970. Six distinctive rock types, one type that was heavily patinated but seems
structurally different from the identifiable groups, and a class composed of unidentifiable
materials (burnt, weathered, etc.) were defined. All lithics were classified according to this
scheme. The six rock types are: volcanic tuff, silicified tuff, cherts, quartz aggregates, opals or
chalcedonies, and rocks of effusive origin. In general, lithic materials at Krapina, both waste
and tools, are dominated by tuffs and silicified tuffs (65%); cherts, comprise around 23% of the
artefacts (Table 1). Frequencies are similar in tool and waste distributions. The overall pattern
      567

Table 1 Frequencies of lithic raw material classes at Krapina

Material Total Waste Tools

M1 Tuffs 0·47 0·48 0·42

M2 Silica tuffs 0·18 0·19 0·16
M3 Chert 0·23 0·22 0·25
M4 Aggregate quartz 0·01 0·01 0·02
M5 Opal 0·01 0·00 0·01
M6 Effusive 0·00 0·00 0·01
M7 Patinated 0·09 0·08 0·01
M8 Unknown 0·01 0·01 0·13
n 1191 918 273

at Krapina seems to suggest that little selection based on raw materials occurred between
initial reduction and tool production. For the present study, only seven rock types are used; the
unidentifiable class (which is small, anyway) is excluded from the comparisons that follow.

Technology
A final aspect of chronological change that might illuminate varying use of the Krapina shelter
by Middle Paleolithic people is technology of tool blank production. As was discussed earlier,
the bulk of Krapina waste materials indicates that a cobble wedge technology was used to
produce blanks for tools, at least at the site-wide assemblage level. One result of this technology
is an abundance of blanks with partial cortex, particularly naturally backed knives (sensu
Bordes). At the same time, Levallois flakes are present, and it is our impression that these are
especially common as blanks for tools made in exotic materials (although sample sizes are too
small to test this impression quantitatively with exotic materials alone).
With these general characteristics of the Krapina stone tool assemblage in mind, we turn
now to chronological analyses designed to identify potential behavioral changes over time.
First, we consider the individual strata distinguished in the field by Gorjanović. Then, we
consider the two grosser units suggested by actual artefact vertical distributions.

Gorjanović’s ten geological strata

Typology
The basic typological character of the overall assemblage is retained in consistent fashion when
Gorjanović’s ten geologic stratum assemblages are considered (Figure 4). Comparison of
typological indices shows that sidescrapers vary between 50 and over 90% of the tools in a
level, and other types vary in relation to the scraper index. The scraper index, as well as those
for other classes, increases and decreases through the stratigraphy. No trend in typological
composition is evident to suggest systematic behavioral change over time, and the lack of
temporal patterning suggests simple stochastic variation in a scraper-dominated Mousterian
assemblage. It is possible, of course, that sample size affects variation in the scraper index,
especially because level sample sizes are so small. However, a regression analysis of the effects
of sample size on scraper frequencies produced an insignificant fit between the distributions;
the regression slope could not be distinguished from 0 and an R2 of 0·0001 emphasizes the lack
of relationship. Thus, fluctuations in scraper frequencies over time cannot be attributed to
568 . .   . . 

0.8

f 0.6

0.4

0.2

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 4. Distribution of various tool indices (Bordes, 1961) across strata. (/) Sidescraper index, (.) Backed
knife index, ( ) Index for Upper Paleolithic tools, and ( ) Notched and Denticulated tool index.

0.6

0.5

0.4

0.3
f

0.2

0.1

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 5. Distribution of volcanic tuffs (M1 in Table 1) across strata.

sample size effects. Essentially, the Mousterian from Krapina remains typologically unchanged
over the entire course of site occupation, even though precise frequencies of its constituent
elements vary randomly.

Raw materials
As Table 1 would suggest, four materials are rare in all levels: quartz aggregates, chalcedony,
effusive rocks, and the patinated (but distinctive) group. According to Zupanič (1970), these are
almost certainly ‘‘exotic’’ materials with probable sources at some distance from the site. Three
materials are abundant in the assemblages: volcanic tuff, silicified tuff, and chert. Collectively,
these materials make up 88% of all the raw materials composing the Krapina assemblage.
Interestingly, they do not seem to co-vary in frequency across the levels. Different individual
materials dominate in different levels.
The distribution of volcanic tuff as a raw material for stone tools is illustrated in Figure 5.
(In this and all subsequent figures, frequencies are calculated as the proportion of the entire
provenienced assemblage from each level composed of a given material type.) Tuffs are by
far the most frequent material used in levels K, 1, 5 and 9. Assessments performed with
Jakov Radovčić in the Krapinica River below Krapina, showed that more than 75% of the
      569

0.4

0.3

0.2
f

0.1

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 6. Distribution of silicified tuffs (M2 in Table 1) across strata.

0.7

0.6

0.5

0.4
f

0.3

0.2

0.1

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 7. Distribution of flints/cherts (M3 in Table 1) across strata.

gravel in the sediment load of the modern river is composed of tuff (cf. Simek, 1991). It is
curious that this most common of local materials dominates only in some of the smaller
assemblages.
Silicified tuffs are most common in levels 3, 5, 7 and 8 (Figure 6). Thus, they tend to be
frequent in at least some levels where volcanic tuffs are least common, and they dominate
more in later levels than in earlier ones. The two levels with concentrations of tools and
hominid fossils, levels 3 and 8, have high frequencies of silicified tuffs. This material comprised
only 1% of the modern Krapinica gravel samples. Thus, either it was procured elsewhere or
it was preferentially selected during procurement at the river.
Chert (Figure 7), is most common in levels 4 and 6 and is relatively rare in other levels.
Thus, cherts tend to occur where silicified tuffs do not. They are not abundant in levels 3 and
8, however, where hominid remains are most common. In the modern river gravels, they
composed 6% of our samples.
Thus, two groups of abundant raw materials seem to be indicated. Volcanic tuffs, by far the
most common raw material in the gravel of the modern Krapinica River below the site, are the
most frequent tool material in early levels at Krapina and also dominate material profiles in
levels 5 and 9. Silicified tuffs and cherts, both generally higher-quality materials, collectively
570 . .   . . 

0.3

0.25

0.2

0.15
f

0.1

0.05

0
K 1 2 3 4 5 6 7 8 9
Stratum
Figure 8. Frequencies of Naturally Backed Knives, i.e., cobble wedges, (/) and Levallois flakes (.) used as
tool blanks across strata.

dominate the later levels at Krapina, but do not co-occur; silicified tuffs are frequent in levels
3, 5, and 8, where cherts are common in levels 4 and 6. Both are relatively rare in the sediment
load of the Krapinica, and must have either been imported into the site with other rare
‘‘exotics’’ procured at some distance, or preferentially chosen from available cobbles at the
river’s edge. Finally, only silicified tuffs dominate in the levels with the richest accumulations
of hominid remains.

Technology
As we have seen, two kinds of debitage blanks dominate the Krapina assemblage, cobble
wedges (typologically, naturally backed knives) and Levallois flakes. It is possible that there is
technological change over time from one of these blank forms to the other. Figure 8 shows the
frequency of naturally backed knives and Levallois blanks for all provenienced pieces by level.
It is clear from this figure that these two production techniques were not used in equal
proportions at any given time. The Levallois technique, while relatively rare overall, was the
only method used in level K, dominates the level 2 assemblage, and is the more-frequent
technique used during the level 8 occupation. Naturally-backed knives are most common in
level 9 at the end of the sequence, and they dominate the middle levels (3, 4, 5, and 7).
Levallois blanks are more common than naturally-backed knives in level 6, the level with the
fewest artefacts overall. Thus, there would seem to be chronological variability in technology
at Krapina, even though the typological character of the assemblage remains unchanged over
time. The relationship between the two technologies (cobble wedge vs. Levallois) is complex,
however, and a regression analysis found no linear relationship at all, either positive or
negative (R2 =0·007). It cannot be said, therefore, that one technology replaces the other.
Indeed, in the hominid-rich levels, a general mix of technologies is evident; naturally-backed
knives are somewhat more frequent in level 3, while Levallois blanks are slightly more
common in level 8. These mixtures, it should be noted, correspond to the richest layers in
terms of artefacts, so some of these variations might be due to sampling effects. At the same
time, larger assemblages might be related to differences in site use that also resulted in the
deposition of hominid remains. At present, it is difficult to distinguish between these
possibilities with the data at hand.
      571

0.6

0.5

0.4

0.3
f

0.2

0.1

0
Lower Upper
Stratum
Figure 9. Comparison of various tool indices (Bordes, 1961) between upper and lower artefact stratigraphic
concentrations (see Figure 6). (/) Sidescraper index, (.) Backed knife index, ( ) Index for Upper Paleolithic
tools, and ( ) Notched and denticulated tool index.

0.5

0.4

0.3
f

0.2

0.1

0
Lower Upper
Stratum
Figure 10. Comparison of various raw material type frequencies between upper and lower artifact
stratigraphic concentrations (see Figure 5). (/) Volcanic tuffs, (.) Silicified tuffs, ( ) Flints/cherts, and
( ) All other (exotic) materials.

Upper and Lower Unit assemblages

Typology
The essential typological consistency observed for the geologic levels is also apparent when
typological indices are compared between the Upper and Lower Units (Figure 9). In both
units, sidescrapers dominate the assemblages by a wide margin; the frequency is slightly less in
the Upper Unit, but the scraper index still exceeds 0·4 in the later assemblage. Notches and
denticulates are slightly more frequent in the Upper Unit, but fail to reach 0·1 in either Unit.
Typological homogeneity is evident at all scales of stratigraphic resolution at Krapina; the
assemblages change little over time in this dimension.

Raw materials
Raw material frequencies differ between the Lower and Upper Units (Figure 10), reflecting
variation in material use seen at the finer stratigraphic scales. The primary difference lies in a
relative increase in the frequency of chert in the Upper Unit, and a coincident decrease of
572 . .   . . 

0.12

0.1

0.08

0.06
f

0.04

0.02

0
Lower Upper
Stratum
Figure 11. Comparison of frequencies of (/) backed knives as tool blanks and ( ) Levallois blanks between
upper and lower artefact concentrations.

Silicified Tuffs. The increase in chert frequency is dramatic enough to affect frequencies for
other material types. This pattern was also observed at the finer stratigraphic scales. Overall,
changing selection of raw materials over time, with the better-quality cherts favored in later
time periods, is suggested. Interestingly, there is no difference between Lower and Upper Units
in other exotic material frequencies.

Technology
Only a single technological change was suggested at the ten-stratum level, and the same trend
is also evident here (Figure 11). Naturally-backed knives, produced by the cobble wedge
exploitation technique outlined earlier (Simek, 1991), are equally common in both Lower and
Upper Units. The Levallois technique, however, while present in both Units, is more frequent
earlier in time. This is generally consistent with technological change during the Levalloisian
Mousterian as seen elsewhere in Europe. At Vaufrey cave, for example, the Levallois
technique dominates Couche VIII, the first Mousterian level at that site, which was deposited
during the Riss glacial either late in Isotope Stage 7 or at the beginning of Stage 6 (e.g.,
Geneste, 1985; Rigaud, 1982).
It is of particular interest that Kuhn (1995), examining a series of Pontinian Mousterian
assemblages from central Italy, observed a similar, but more-pronounced, shift in technology
from Levallois flake production to cobble wedges (naturally-backed knives). According to
Kuhn, this shift occurred about 50,000 years ago in Italy, and he suggests that it may reflect
changing technology at the end of the Mousterian or beginning of the Upper Paleolithic
linked to evolving land use patterns. The cobble wedge technique might, at least superfici-
ally, resemble a kind of incipient blade technology. However, in Italy and at Krapina, the
technique is linked to exploitation of raw materials in small packages, i.e., river cobbles. While
it is true that a number of Italian industries produced on small cobbles show the use
of the Levallois technique (Kuhn, 1995; Simek & Ammerman, 1990), the shift to cobble
wedging might simply reflect changing technological adaptations to the material itself, rather
than a wider process of biocultural change. This interpretation is supported: (1) by a lack
of real technological replacement at Krapina, and (2) by the dating of Krapina, which
would place any technological changes at more than 100,000 years ago, much earlier than
in Italy.
      573

Conclusions

Overall, analyses at both fine and coarse stratigraphic scales indicate similar patterns of change
and stasis over time through the Krapina sequence. The Krapina site was occupied exclusively
by Mousterian tool makers at least twice and possibly more times. While there, raw materials
obtained in nearby river gravels were made into tools by opening and sectioning naturally-
occurring cobbles to produce flakes. Sidescrapers are consistently the most common tool
group, although various other Mousterian tool forms are present. Some tools were brought
into the site in finished form as part of a curated technology. Other tools were expediently
produced on a wide range of flake forms, using nearly all the products of the reduction
sequence.
When the provenienced artefacts are divided into the ten stratigraphic assemblages defined
by Gorjanović during his excavations, some change in assemblage characteristics can be
observed. Considering the small sample sizes involved and the nature of the excavations, the
level assemblages show remarkable typological consistency. All levels have scraper-rich
(Charentian) Mousterian assemblages, with variation in the scraper index among levels never
dropping below 50% of the typed tools. What variation does exist among level assemblages lies
in raw material procurement and technology. In early levels, river cobbles were brought up to
the site that essentially reflected material proportions in the parent gravels (i.e., volcanic tuffs
dominate the raw materials used for tools). This suggests little selection of material for use. The
Levallois technique was used frequently in some of these early levels, with its greatest
frequency occurring in level 2. In later levels, cobbles were selected of higher-quality cherts
and silicified tuffs; these were exploited using the wedge technique first described by Baumler.
The levels with the richest assemblages and hominid concentrations, levels 3 and 8, exhibit a
mixed technology with both Levallois blanks and backed knives present in similar abundance.
Silicified tuffs were the preferred raw materials in both hominid-bearing levels, and there is
little else to suggest important differences in site use between these two layers. Still, the
technological changes we observe represent important shifts over time at Krapina: (1) later site
occupants selected better quality cherts to produce their tools, and (2) they used a more
efficient means for exploiting the cobble ‘‘packages’’ of material available to them in the
sediment load of the river below the site. When assemblages are grouped into maximal
stratigraphic units based on vertical frequencies, similar patterns are indicated. Thus, the
observed tendencies, both for change and stasis, are indicated regardless of stratigraphic scale.
Interpretively, the Krapina assemblages seem to reflect occupations by rather mobile groups
using lithic resources in complex fashion. Mobility is indicated by the small assemblage sizes in
all levels, suggesting brief occupations, and by the presence of exotic stone raw materials in all
levels (Simek, 1991). Layers with few or no hominid remains (and few artefacts) can vary
significantly in raw material selection and technology, although (as we noted above) there are
general changes in these aspects of the assemblages over time. The hominid-bearing deposits
are similar to each other, although they differ from the less-rich layers in various dimensions.
While no real evidence for long-term occupation is present in any layer, varying patterns of site
use over time do seem to occur within the context of assemblages that are in essential typology
unity.
As was discussed earlier, the Krapina 1 or ‘‘A’’ cranium from level 8 and other specimens
from the Upper Unit may exhibit somewhat more modern characteristics than appear in the
earlier fossils from levels 3/4. If this is the case, then it may be that changes seen in certain
aspects of the artefact record, i.e., in-site use and regional resource procurement, may parallel
574 . .   . . 

biological changes in local Neandertal populations. On the other hand, if the morphological
pattern of Krapina 1 and other Upper Unit remains cannot be systematically differentiated
from other Neandertals, then the observed behavioral changes occur independent of
significant biological change. In any case, the behavioral context in which the hominid
remains from Krapina were deposited is fundamentally the same for both the Lower and
Upper Units. We probably cannot go much further than this in our interpretations. Still,
because Gorjanović took great care in his excavations and kept nearly all the stone artefacts he
found, we have been afforded a more detailed view of the Mousterian occupations than is
usually possible from data collected at the turn of the century.

Acknowledgements
We would like to express our thanks to Jakov Radovčić, Marina S {imek, and the late Mirko
Malez, along with the University of Tennessee Research Development program, for making
studies of the Krapina stone tools possible. We are also grateful to Nancy Minugh-Purvis and
to the Wenner Gren Foundation, the National Academy of Sciences, and the Alexander Von
Humboldt Foundation for supporting studies on the Krapina hominids.

References
Bar-Yosef, O. & Pilbeam, D. (1993). Dating hominid remains. Nature 366, 415.
Baumler, M. (1988). Core reduction, flake production, and the Middle Paleolithic industry of Zobiste (Yugoslavia). In
(H. Dibble & P. Mellars, Eds) Upper Pleistocene Prehistory of Western Eurasia, pp. 255–274. Philadelphia: University of
Pennsylvania Press.
Bordes, F. (1954). Les limons quaternaires du bassin de la Seine. Archives de l’Institut de Paléontologie Humaine Memoire 26.
Paris: IPH.
Bordes, F. (1961). Typologie de Paléolithique Ancien et Moyen. Bordeaux: Delmas.
Caspari, R. (1991). The evolution of the posterior cranial vault in the central European Upper Pleistocene. Ph.D.
Dissertation, University of Michigan.
Frayer, D. (1992). Evolution at the European edge: Neanderthal and Upper Paleolithic relationships. Préhist. Europ. 2,
9–69.
Geneste, J.-M. (1985). Analyse Lithique des Industries Mousteriennes du Périgord: Une Approache Technologique du
Comportement des Groupes Humains au Paléolithique Moyen. Ph.D. Dissertation, Université de Bordeaux I.
Gorjanović-Kramberger, K. (1899). Der paläolitische Mensch und seine Zeitgenossen aus dem Diluvium von Krapina
in Croatien. Mitt. anthrop. Ges. Wein 29, 65–68.
Gorjanović-Kramberger, K. (1900). Der diluviale Mensch aus Krapina in Kroatien. Mitt. anthrop. Ges. Wein 30, 203.
Gorjanović-Kramberger, K. (1901). Der paläolithische Mensch und seine Zeitgenossen aus dem Diluvium von
Krapina in Kroatien. Mitt. anthrop. Ges. Wein 31, 164–197.
Gorjanović-Kramberger, K. (1902). Der paläolithische Mensch und seine Zeitgenossen aus dem Diluvium von
Krapina in Kroatien. Nachtrag (als zweiter Teil). Mitt. anthrop. Ges. Wein 32, 189–216.
Gorjanović-Kramberger, K. (1904). Der paläolithische Mensch und seine Zeitgenossen aus dem Diluvium von
Krapina in Kroatien. Zweiter nachtrag (als dritter Teil). Mitt. anthrop. Ges. Wein 34, 187–199.
Gorjanović-Kramberger, K. (1906). Der diluviale Mensch von Krapina in Kroatien. Ein Beitrag zur Paläoanthropologie.
Weisbaden: Kreidels Verlag.
Gorjanović-Kramberger, K. (1913). Život i kultura diluvijalonga čovjeka iz Krapine u Hrvatskoj. Djela jugosl. Akad.
Znan. Umjetn. 23, 1–54.
Guenther, E. (1959). Zur Altersdatierung der diluvialen Fundstelle von Krapina in Kroatien. Bericht über die 6.
Tagung der Deutschen Gesellschaft für Anthropologie 202–209.
Hrdlicka, A. (1930). The skeletal remains of early man. Smithsonian Miscellaneous Collections 83.
Klaatsch, H. & Hauser, O. (1910). Homo aurignacensis Hauseri, ein paläolithischer Skelettfund aus dem unteren
Aurignacien der Station Combe-Capelle bei Montferrand (Perigord). Präehist. Z. 1, 273–338.
Kuhn, S. (1995). Mousterian Lithic Technology. Princeton: Princeton University Press.
Malez, M. (1970a). Rezultati revizije pleistocenske faune iz Krapine. In (M. Malez, Ed.) Krapina 1899–1969,
pp. 45–56. Zagreb: Mladost.
      575

Malez, M. (1970b). Novi pogledi na stratigrafiju krapinskog nalazišta. In (M. Malez, Ed.) Krapina 1899–1969,
pp. 13–44. Zagreb: Mladost.
Malez, M. (1970c). Paleolitska kultura Krapine u svjetlu novijih istraživanja. In (M. Malez, Ed.) Krapina 1899–1969,
pp. 57–129. Zagreb: Mladost.
Malez, M. (1978). Stratigrafski, paleofaunski, i paleolitski odnosi krapinskog nalazišta. In (M. Malez, Ed.) Krapinski
Pračovjek i Evolucija Hominida, pp. 61–102. Zagreb: Mladost.
Minugh-Purvis, N. (1988). Patterns of craniofacial growth and development in Upper Pleistocene hominids. Ph.D.
Dissertation, University of Pennsylvania.
Minugh-Purvis, N. (1997). Variation in immature hominid neurocranial morphology at 100–130 ka: a comparison of
the es-Skhul and Krapina 1 juveniles. In (T. Akazawa & O. Bar-Yosef, Eds) Neanderthals and Modern Humans in West
Asia. New York: Plenum Press. Museum. (In press).
Minugh-Purvis, N., Radovčić, J. & Smith, F. (1996). The place of the Krapina 1 juvenile in late Pleistocene hominid
evolution. Am. J. phys. Anthrop. 22 (Suppl.), 170.
Radovčić, J. (1988). Gorjanović-Kramberger and Krapina Early Man. Zagreb: Skolska Knjiga/Hrvatski Prirodoslovni Muzej.
Radovčić, J., Smith, F. H., Trinkaus, E. & Wolpoff, M. H. (1988). The Krapina Hominids: An Illustrated Catalog of Skeletal
Collection. Zagreb: Mladost.
Rigaud, J-Ph. (1982). Le Paléolithique en Périgord: Les Donnes du Sud-ouest Sarladais et leurs Implications. Ph.D.
Dissertation, Université de Bordeaux I.
Rink, W. H., Schwarcz, H., Smith, F. & Radovčić, J. (1995). ESR ages for Krapina hominids. Nature 378, 24.
Schwalbe, G. (1904). Die Vorgeschichte des Menschen. Stuttgart: Schweizerbart.
Simek, J. (1991). Stone tool assemblages from Krapina (Croatia, Yugoslavia). In (A. Montet-White & S. Holen, Eds)
Raw Material Economies among Prehistoric Hunter Gatherers, pp. 59–72. University of Kansas Publications in Anthro-
pology 19. Lawrence: University of Kansas Press.
Simek, J. & Ammerman, A. (1990). Lithic raw material use in the Mousterian of Burrone Scierra I (Calabria, Italy).
In (M. Lenior, Ed.) Le Silex de sa Genese a l’Outil, pp. 479–488. Cahiers du Quaternaire No. 17. Bordeaux: C.N.R.S.
{kerlj, B. (1958). Were Neanderthalers the only inhabitants of Krapina? Bull. scien. Cons. Acads. RPF Yougosl. 4, 44.
S
Smith, F. H. (1976). The Neandertal Remains from Krapina: A Descriptive and Comparative Study. University of Tennessee
Reports of Investigations 15. Knoxville: University of Tennessee Press.
Smith, F. H. (1982). Upper Pleistocene hominid evolution in south-central Europe: a review of the evidence and
analysis of trends. Curr. Anthropol. 23, 667–703.
Smith, F. (1997). Dragutin (Karl) Gorjanović-Kramberger. In (F. Spencer, Ed.) Encyclopedia of the History of Physical
Anthropology. New York: Garland.
Trinkaus, E. (1977). A functional interpretation of the axillary border of the scapula. J. hum. Evol. 6, 231–234.
Trinkaus, E. (1978). Functional implications of the Krapina neandertal lower limb remains. In (M. Malez, Ed.)
Krapinski Pračovjek i Evolucija Hominida, pp. 155–192. Zagreb: Mladost.
Turq, A. (1992). Raw material and technological studies of the Quina Mousterian in Perigord. In (H. Dibble &
P. Mellars, Eds) The Middle Paleolithic: Adaptation, Behavior, and Variability, pp. 75–86. Philadelphia: University of
Pennsylvania Press.
Wolpoff, M. (1979). The Krapina dental remains. Am. J. phys. Anthrop. 50, 67–114.
Wolpoff, M. (1980). Paleoanthropology. New York: Knopf.
Wolpoff, M. (1995). Human Evolution. New York: McGraw-Hill.
Zupanič, J. (1970). Petrografska istraživanja paleolitskih artefakta krapinskog nalazišta. In (M. Malez, Ed.) Krapina
1899–1969, pp. 131–140. Zagreb: Mladost.