Historiography of the issue
From the moment of its appearance on the archaeological scene to the present, the Kostenkov-type recess tip is a shape that is defined by a wide range of characteristics that reflect the archaeological era in its specific cultural refraction. The finality of the tip shape, its striking features and repeatability in the absence of obvious standardization make researchers turn to different aspects of its morphology, depending on the scientific concept that they share.
At the first stages of studying the tips with a notch, one can see two mutually exclusive typological approaches proposed by P. P. Efimenko and M. D. Gvozdover. Thus, P. P. Efimenko defined arrowheads as the most typical tools of the Aurignacian-Solutrean era, continuing the tradition of Solutrean leaf-shaped points [1958, pp. 233-234]. The researcher essentially did not single out the Kostenok-1 tips as a special type. He saw in the "blade-tip" a pan-European category, in which the features of the progressive development of plate technology were revealed: a plate base with precise and fine adjustment-this is the essence of the Kostenkov form. Possible asymmetry of the blank was corrected by the asymmetry of the petiole. According to the researcher, the pen corrected with a thin flat retouch corresponded to its functional use in complicated labor operations. He gave the following typological characteristics of blade tips: 1) a regular or asymmetric plate that does not need to completely change its shape, of three main sizes: large-10 cm, medium-5.5 cm, small-5.5-2.5 cm; 2) has a notch forming a petiole; 3) with pointed (to varying degrees)teeth. the lower and upper ends of the tip; 4) the plate is thinned with a flat Solutrean retouch; 5) there are no signs of dimensional standardization.
The proposed features described the tool at the level of the category, for which the point and recess on a plate base were essential. Solutreyan retouching carried an "epoch-making" load. According to the author, all other signs were not dominant.
The point of view of P. P. Efimenko presented here was related to the final phase of the stadium typology of the 1930s-1950s. By the time of publication, ideas about the local diversity of the Upper Paleolithic were already being actively formed. The leading forms of flint tools became the basis for determining the local and even "ethnographic" features of the Avdiivka site by M. D. Gvozdover. She emphasized the specific features of the tool, its differences from others, primarily Solutrean and Madeleine [1961, p. 112-115; Gvozdover and Rogachev, 1969, p. 487-500]. The task of identifying the characteristic shape of the tip led the researcher to identify a group of so-called atypical tips - small, morphologically very simple. The presence of such tools in the group of" Kostenkov tips " is rare.-
page 16
it took an already difficult form to describe (message in a personal conversation, 2004).
According to M. D. Gvozdover, the "Kostenkov type tip" has the following morphological features: it is made on regular wide plates; the length is 12-14 cm; the recess is highlighted by a steep retouch and occupies 2/3 of the length of the tool; the pen and base on both sides of the tip are worked out by Solutrean retouching.
The list of attributes has become more strict and shorter-without "atypical" forms. The inclusion of quantitative and qualitative characteristics allowed us to see the differences between Kostenkov and Solutreysky tips, even with the remaining signs of Solutreysky retouching in the list. According to the selected features, the Kostenkov tip is a large, rather massive tool with a shortened broad feather. The description made by M. D. Gvozdover determined for decades the status of the Kostenkov tip as a special typological form. It was then that the tip was classified as a type. This description has become the basis for modern European classifications [Demars and Laurent, 2000, p. 138]. It serves as a sufficient criterion for assigning collections from newly discovered sites to the Kostenkov circle (Trusov, 1998, p. 291).
Long-term fieldwork at the new sites of Kostenok-1 and Avdeevo in 1970-1990, the discovery and study of the brilliant Zaraiskaya site made it necessary to compile a complete typological description of the leading East Ossetian industries. Their flint collections included, along with classical samples, numerous forms that have only some features of a typical Kostenkov tip. The petiole was sometimes slightly less than half the length of the tool, the feather could be long and narrowed; the main proportions varied; the size of the tips ranged from 10-12 to 2.5 cm; signs of flat ventral retouching at the ends could be absent. The classical type, therefore, turned out to be one of the variants of the Kostenkov tip. It "worked" only in the external space of analysis, marking monuments of the Kostenkov culture, but did not help to differentiate even within the limits of the Kostenkov-Willendorf unity. It was necessary to understand the peculiarities of the formation of tools within the same culture or even one parking lot. The morphological diversity of the tips of the Kostenkov sites completely did not fit into the framework of one specific Kostenkov type.
The difficulties of identifying the tips forced researchers to turn to descriptive analysis with elements of typological reconstructions. The previously formed view of the unity of the tip type, however, was not destroyed. However, the type as a unit of archaeological classification no longer seems to be something rigid and unambiguous. There is an idea of several levels of typological stability. Moreover, these levels seemed clearer to some researchers, and more variable to others, depending on a number of factors. This approach gave an impetus to the description of the category at different typological levels. This line of research is most clearly expressed by M. D. Gvozdover when describing two sites of the Avdiivka parking lot [1998, p. 259-263]. The signs of the tips were divided into mandatory and optional ones. The following attributes are selected as mandatory:: 1) the presence of a notch that makes up 2/3 of the tip length; 2) traces of steep retouching on the notch; 3) a sharp transition from the notch to the pen; 4) the pen is triangular, symmetrical; 5) the pen angle is stable; 6) the edge opposite to the notch is convex, arched. They also probably include the size of the tips, which are divided into three groups in length: large and medium - from 5.5 to 12 cm, small - from 2 to 4.5 cm.
Mandatory attributes allowed the author to include all the tips in a single array and abandon the concept of"atypical tip". Indeed, if we attribute to atypical tips that do not have traces of flat ventral retouching, then their group will significantly expand due to large tips. Atypical forms in this case can be considered some large tips Kostenok-1 and probably many tools Zaraiskoy parking.
Signs of secondary treatment were determined by M. D. Gvozdover as optional or accidental. Describing each group of tips, she expressed deep and probably correct assumptions about the causes of the manifestation of certain facultative, mostly descriptive features. However, there are still doubts about the clear distinction between optional and mandatory features. Artifacts with a shortened or modified pen, or a modified arc-shaped edge - modifications of typical tips.
Typologization of large arrowheads from the Zaraisk site is also difficult due to the lack of clear ideas about standard forms and methods of secondary processing. Kh. A. Amirkhanov [1998, p. 24; 2000, p. 207-208] and after him S. Yu.Lev [2003] give descriptive characteristics of arrowheads. Their generalizations were made at the category level. Typological features include a notch on the petiole, which, however, does not have a standard shape. Other characteristics can be attributed to the category of the same optional, variable, not constant, but very specific. -
page 17
However, this is only a general statement, and specific descriptions are filled with various, changing features [1998, p. 279-298].
Summing up the above, we note that the main result of many years of studying the tips of the East Gravettian sites was the determination of a narrow specific group that qualifies as a type. At the same time, the variety of tips is well known, many of which have only individual characteristics of the type. Was there a variety of types, or did the same type of weapon change? In the latter case, the classification unit includes a wide range of morphological features. In this regard, it is important to assess the degree of shape variability at the level of both one monument and a group of related industries.
As follows from general theoretical provisions, the variability of any type of product can be caused by various factors - randomness, purposeful actions in the process of manufacturing products, as well as subsequent modifications (adjustments) during operation. The most likely complex effect on the formation of the final morphology of the product. In our opinion, identifying the specific role of each of the above factors is an almost unsolvable task when using only traditional descriptive methods of typological analysis due to the lack of criteria for assessing the boundaries of random and regular. The proposed approach is based on statistical analysis of data from several measurements that characterize the main morphological features of Kostenkov tips, as well as the specifics of their secondary processing. The analysis of measurement and quality characteristics using one-and multidimensional statistics, in our opinion, will allow us to estimate the observed tip morphology in terms of not only statics (taking into account its final shape), but also dynamics (from the point of view of the process that could lead to this shape).
So, the purpose of this paper is to evaluate the variability of Kostenkov tips and reconstruct the possible causes of this variability.
Research material and program
The sample consists of 138 whole tips with a recess of two Kostenok-1 complexes (upper layer). The old complex is represented by the collection of P. P. Efimenko (70 copies), the new one-by the collection from excavations in the 1970s-1980s (68 copies).
The final description program for the Kostenkov tips was developed by V. I. Belyaeva (1979; Rogachev et al., 1982, pp. 50-56). She also made all the measurements (Table 1).
To describe the morphology, the following reference points were fixed on the tip (Fig. 1): T and B are the extreme points of intersection of the tip contour with the upper and lower horizontal tangent lines, respectively;
L and L' are the points at the intersection of the upper and lower horizontal tangent lines and the tangent to the side opposite the recess; W and H are the extreme points of intersection of the tip contour with the right and left sides.
Table 1. Values of measuring features of the tips of the Kostenka parking lot-1
Code
Type of retouching
Length, mm
Width, mm
Pen length, mm
Arc
Angle, deg.
Oval
on the back
on the abdomen
1
2
3
4
5
6
7
8
9
New complex
1
But
0
82
18
32
10
31
7
2
But
X
73
14
25
9
32
31
3
But
0
74
25
26
14
50
26
4
But
0
63
20
19
11
54
15
8
But
B
62
17
15
12
61
20
9
X
X
61
18
14
11
62
15
10
0
0*
70
19
23
13
45
19
11
But
0
77
19
26
15
40
12,5
12
But
0
77
20
21
14
48
13
13
X
B
57
20
22
11
50
30
14
X
X
54
17
14
11
64
24
page 18
Continuation of Table 1
1
2
3
4
5
6
7
8
9
15
X
B
61
20
14
12
71
24
16
X
B
70
15
8
19
93
23
17
But
B
63
17
28
10
37
21
18
But
X
60
16
19
9
43
14
19
X
X
61
15
31
9
27
15
20
0
0
9
13
18
10
36
21
21
0
X
52
11
19
7
30
10
22
But
B
50
11
19
6
32
14
23
0
0
54
13
23
7
33
14
24
But
0
40
14
17
10
43
17
25
But
0
38
12
12
9
50
12
26
But
0
32
13
12
8
60
11
27
But
0
42
9
8
6
50
9
28
But
0
38
9
11
4
43
6
30
0
0
42
12
17
5
30
15
31
But
0
31
9
11
5
40
7
32
But
0
32
8
8
5
34
6
33
But
0
32
7
7
5
43
9
34
0
B
57
12
17
7
37
16
64
But
0
63
22
25
12
50
22
70
0
0*
63
26
31
10
43
21
72
But*
B
71
22
24
15
47
28
73
But
0
35
10
11
7
50
9
1k
But
0
58
14
20
7
35
7
3k
X
0
55
15
13
10
61
13
4k
0
0
78
21
21
14
50
13
5k
But
0*
54
15
13
9
58
13
6k
But
0
82
18
33
11
30
15
7k
But
0
78
20
23
15
45
17
8k
But
0
72
15
25
9
33
11
9k
0
B
71
22
20
12
56
20
10k
But
0
57
18
23
10
40
10
11 k
But
B
55
15
14
9
58
14
12k
0
0
53
11
19
6
36
11
13k
But
0
42
9
8
5
66
8
14k
But
0
51
12
19
7
34
12
15k
But
B
65
18
20
11
40
21
17k
But
B
78
20
20
14
55
11
18k
0
0
79
22
28
14
41
19
19k
But
B
89
23
24
14
47
17
20k
But
0
78
17
24
8,5
38
7
21k
But
0
73
18
24
12
40
12
22k
0
B
58
20
18
14
59
22
page 19
Continuation of Table 1
1
2
3
4
5
6
7
8
9
25k
0
0
50
13
17
9
43
10
27k
But
0
66
19
20
12,5
50
20
29k
But
0
85
20
34
12
32
17
30k
But
B
68
22
16
15
65
24
31k
But
0
58
14
20
10
38
14,5
34k
0
0
39
14
15
8
49
13
35k
But
0
68
18
16
12
50
17
36k
But
0
35
7
9
5
40
8
37k
0
0
39
13
14
8
52
9
38k
X
B
76
21
25
15
47
19
39k
But
B
61
17
15
10
60
12
40k
But
X
56
15
17
9
43
26
41k
But
X
64
16
16
9
52
13
42k
But
0
55
15
23
8
36
5
Old complex
1e
0
X
55
15
20
8,5
40
10
2e
But
B
42
11
13
7
48
10
3e
But
B
65
19
22
13
50
15
4e
But
0
39
11
14
6
42
9
5e
X
0
77
23
34
11
40
17,5
6e
But
X
59
17
13
13
80
10
7e
X
B
63
19
13
13,4
70
21
8e
But
B
69
26
27
10
50
19
9e
But
B
78
25
18
17
70
20
10e
But
*
55
10
9
10
60
13
11e
But
B
89
30
35
20
46
34
12e
But
B
51
17
13
10,5
62
15
13e
But
B
41
11
13
7
48
12
14e
0
0
70
16
38
9,5
26
21
15e
But
0
51
16
18
8
46
11
16e
But
B
71
20
20
17,5
63
20
17e
0
B
62
17
12
14
69
23
18
But
B*
80
23
27
17
48
44
19e
But
X
60
20
17
12,5
60
17
20e
0
B
50
21
15
11,5
62
14
21e
But
B
69
19
13
16
65
14
22e
X
X
55
19
20
11
50
13
23e
But
B
52
15
13
13
65
12
24e
But
B
57
12
20
7
35
16
25e
But
B
57
17
14
18
60
20
26e
But
B
56
14
13
10,5
53
14
29e
But
0
60
18
26
10
40
12
30e
But
0
95
24
37
12
38
24
page 20
End of Table 1
1
2
3
4
5
6
7
8
9
31e
0
B
62
12
9
11,5
72
18
32e
But
B
60
16
15
8
65
16,5
33e
But
B
43
12
11
7,5
55
17,5
34e
X
X
48
13
23
11
30
8
35e
0
0
87
21
27
8
42
25
36e
0
0
40
10
17
8
38
10
37e
0
X
50
9
17
4,3
30
9
38e
But
0
54
14
22
8
38
11
39e
But
0
57
13
27
6
35
8
40e
But
0
50
19
22
11
46
18
41e
0
0
53
14
24
4,5
32
14
42e
But
0
51
17
25
9,3
36
12
43e
But
0
47
12
18
5,5
37
4,5
45e
0
0
35
14
14
7
55
8
46e
But
0
51
14
15
10
50
15
49e
But
0
60
18
20
8
42
7
50e
0
B
53
16
21
8
45
10
51e
But
0
65
20
23
8
48
19
52e
But
0
67
20
23
10
48
23
53e
But
0
43
12
10
8
58
13
54e
0
0
43
15
18
10
45
15
55e
But
0
26
8
7
4
60
7,3
57e
But
0
32
10
10
4,5
57
8
59e
But
X
34
17
11
10,5
53
15
60e
0
0
41
11
12
7
48
12
62e
0
0
40
12
13
6
44
9
63e
0
B
42
9
15
4
35
11
64e
But
0
40
8
12
5
40
14
65e
But
0
60
20
28
7,5
42
22
66e
But
0
51
13
21
4,3
32
8
67e
But
0
57
20
26
7,5
42
15
68e
But
0
34
11
10
7
59
7
69e
0
X
28
7
10
4,2
37
4
70e
But
0
63
15
16
8
50
25
71e
0
0
62
16
27
8,5
33
17
72e
0
0
51
21
30
10,5
45
22
73e
0
0
62
15
28
9
32
16
74e
0
0
45
10
25
6
25
10
76e
But
B
60
17
12
15
68
22
77e
X
X
47
16
13
9
70
4,5
78e
But
0
68
19
17
9
60
7
79e
But
B
43
12
11
8
50
18
* Products with signs of non-manual pen handling (incisive chipping, flat chipping, etc.).
page 21
Fig. 1. Main dimensional characteristics of the tip.
2. Traces of edge (A) and flat (B) retouching.
left vertical tangent lines, respectively; W - projection of the point W on the opposite side of the tip; W" - projection of the point W on the tangent line L - L'; O - point of change of the pen contour by 10° or more relative to the vertical tangent line L - L'; O' - projection of the point O on the upper tangent line; S - point of change or "fracture" of the petiole base contour. The point can be above or on the lower tangent line to the tip contour. H' is the intersection point of the height dropped on the straight line T-S from point H.
Taking into account the presented scheme, we determined the following dimensional characteristics of the tips (the abbreviated name of the features used later in the text is given in parentheses)::
- tip length (length) - distance between points T and B or L and L';
- tip width (width) - distance between points W and W";
- pen tip length (pen length) - distance between L and W points";
- pen tip angle (angle) - angle a between the lines W-T and T-W';
- measure of the curvature of the pen (oval) - calculated as the ratio of the value of the segment O-O' to the value of the segment L-W";
- the value of the arc-shaped curvature of the tip edge opposite to the recess (arc) is calculated as the ratio of the value of the segment H-H' to the value of the segment T-S.
The nature of the secondary processing of the tips was determined by the traces of two retouching options:
1. Type A retouching includes marginal dorsal part-working of the pen, which does not go to the surface and changes only the contour of the edge. Such retouching can be either steep, close to retouching on the recess of the petiole, or semi-circular or rather flat (Fig. 2). In some cases, the signs of edge retouching follow without interruption - from the recess of the petiole to the pen, as if continuing to create the edge contour. The pen outline on the opposite side was processed with the same steep retouching only in isolated cases, which is not enough to get reliable statistical results. If the edge retouching traces are indistinct (a small number of poorly organized facets), it was defined as X.
2. Flat retouching of type B was applied on the ventral part of the pen and on the ventral surface of the petiole. In this paper, we analyzed only the nature of the processing of the pen tip. Flat oval facets cover the surface of the pen at the tip and descend lower along its edges. The upper edge of the pen, which is opposite to the notch, was always processed, while the other edge could remain unprocessed. According to M. D. Gvozdover, this type of retouching may have been used initially when creating the tip "on large massive plates, when the shape of the plate and its irregularities required correction" [1998, p. 259, 263]. You should also pay attention to the fact that the signs of such retouching are especially indicative for reduced tips with a shortened pen. In cases where there are one or two small facets on the ventral part of the feather,
page 22
Processing was evaluated as X. The absence of any traces of retouching was marked with an "O" sign. In some cases, the retouching was supplemented with narrow incisor chips in the form of a dihedral incisor. The reduced areas of the pen could bear the negatives of flattening chips like Kostenkov knives. These secondary features were not taken into account in our analysis; they are not numerous and could not be used in statistical calculations.
We have recorded a number of artifacts that reflect specific processing options for the pen tip:
option
ex.
items with signs of non-manual pen handling (incisive chipping, flat chipping, etc.)
5
ao
53
ah
7
oh
4
XX
6
00
22
AB
27
0V
8
XB
5
X0
1
Reliably detectable treatment options for both sides of the tip tip - AO, 00 and AB-are shown in 102 specimens, or 74 % of the total sample. Next, we divided the sample into four fractions based on pen processing options (Table 2). As a leading indicator, we considered the signs of retouching B, since, as mentioned above, it is usually considered the formative one for Kostenkov tips. For this reason, we combined the artefacts of the AB and AB variants into a single fraction of tips of type B. We also included tips processed according to the CB variant, since it is essentially a difficult-to-determine variant of AB or AB. Tips that do not have traces of retouching B are treated differently in the literature. The criterion is usually the shape features, which are ultimately determined by the dimensional characteristics. If such tips are close to the products retouched by Type B, they are included in the group of classic Kostenkov tips or separated into a separate category of so-called atypical tips.
The artefacts of only the AO variant are assigned to fraction A; artifacts representing the AX variant, which may be a hidden AB variant, should therefore be included in group B. In this case, they are included in the fraction of undetectable artefacts. Another group consists of products with an untreated pen.
Table 2. Main tip fractions selected by pen retouching options
Fraction
Pen processing options
But
ao
B
0B, AB, HB
0 (unprocessed)
00
X (undefined)
AH, XX
All individual data of the sample of Kostenok-1 handpieces according to the qualitative and quantitative indicators described above are shown in Table 1.
Results of statistical analysis
We will try to find out whether the proposed morphological scheme of description has sufficient statistical grounds, as well as to determine the features of shaping associated with the presence of signs of a particular retouching in the product.
At the first stage, we will find out whether there are differences between the tips of fractions A and B in the main metric indicators. Since the number of items in both groups is sufficient to apply parametric methods, the t-test (Student's test) was used. The data in Table 3 show that the fractions differ significantly in four out of six features. For two indicators - tip length and pen length - the differences do not reach the confidence level. In general, the tips of fraction B are larger. This characteristic is of great importance for solving the problem of the ratio of these fractions.
To establish the relationship between the initial six measurement features in two groups of handpieces, a correlation method was used. As shown in Table 1. 4 and 5, most features in both cases are associated with a fairly high and statistically significant correlation, and the coefficients are generally similar in both groups of products. It is characteristic that the strongest connections reflect the features of the technology of manufacturing blanks. For example, the main dimensions of the tip, despite the likely further shape change, were undoubtedly determined by the dimensions of the initial blank, the length and width of which are technologically interrelated indicators, which is reflected by the high value of the correlation coefficient between them (r = 0.84 for fraction A and r = 0.81 for fraction B). The correlation coefficients between most of the indicators have a positive sign, i.e. an increase in the values of one of them is accompanied by an increase in the values of other indicators, which means that
page 23
Table 3. Average values of metric features of tip fractions A and B and values of the t-criterion
Sign
M A
M B
NA
N B
SDA
SD B
t
d.f.
p
Length
55,4
61,2
53
40
16,8
11,7
-1,97
91
NS
Width
15,1
17,5
53
40
4,6
4,6
-2,49
91
<0,05
Pen length
19,1
17,2
53
40
7,5
5,7
1,38
91
NS
Arc
8,60
11,8
53
40
2,9
3,8
-4,62
91
<0,001
Corner
44,1
55,5
53
40
8,7
12,7
-5,15
91
<0,001
Oval
12,5
17,7
53
40
5,3
5,3
-4,69
91
<0,001
Note. M is the arithmetic mean, N is the number of the corresponding tip fraction, SD is the standard deviation, t is the value of the t-criterion, d. f. is the degree of freedom, p is the significance level, and NS are not significant differences.
Table 4. Pairwise correlation coefficients between the initial features for the tip fraction A
Sign
Length
Width
Pen length
Arc
Corner
Oval
Length
1,00
Width
0,84
1,00
Pen length
0,86
0,82
1,00
Arc
0,76
0,82
0,62
1,00
Corner
-0,34
-0,10
-0,55
-0,04
1,00
Oval
0,52
0,70
0,50
0,60
0,06
1,00
Note. Here and further, statistically significant correlation coefficients (p<0.05) are highlighted in bold.
See Table 5. Pairwise correlation coefficients between the initial features for the tip fraction B
Sign
Length
Width
Pen length
Arc
Corner
Oval
Length
1,00
Width
0,81
1,00
Pen length
0,60
0,65
1,00
Arc
0,74
0,70
0,23
1,00
Corner
0,15
0,16
-0,56
0,53
1,00
Oval
0,52
0,54
0,44
0,56
0,10
1,00
Note. Statistically significant correlation coefficients (p < 0.05) are shown in bold.
This indicates the great role of the parameters of the plate shape of the workpiece in determining the final shape of the product. A high and significant negative correlation coefficient is observed only between the pen length and its internal angle. As the pen length decreases, its internal angle increases. In terms of size, this relationship is not particularly large, but technologically it is quite significant, since it may indicate that the tip is being used part-time. The relationship between these indicators in both samples is almost identical (r = -0.55 for fraction A and r = -0.56 for fraction B). This may indicate that both types of retouching are used to correct tools to the same extent. Interestingly, this process did not have much effect on the overall tip length; this is reflected in the low correlation coefficients of these indicators in both groups. The biggest difference in the coefficients of pairwise correlations in the two groups is observed between the angle of the pen tip and its du-
page 24
goy. If there is no such relationship in fraction A, then in fraction B there is a moderate positive relationship of these indicators. An increase in the tip edge bulge is associated with a decrease in its length at a stable arc height. If the initial parameters were similar, its length and tip length should decrease with increasing pen angle without changing the height of the edge arc, but with an increase in its relative value. The lack of correlation between the length of the tip and the angle of its pen does not contradict the above, but suggests initially different lengths of the tips, which will later change the angle. It can be assumed that there is no rigid standard in the initial size of the tip and its blank. For tips with traces of retouching A, there is no such pattern, i.e. changing the angle of the pen did not affect the length of the tool. Thus, although the two groups of handpieces under consideration have significant differences in most metric indicators, their intra-group relationships are similar, and therefore, the technologies of their production and further use were also similar. At the same time, the parameters of the initial blank played a leading role in shaping.
Let's turn to the possibilities of multidimensional statistics. To analyze metric data, we used the principal component method (GC). Its essence consists in moving from the original to new independent features, which are a combination of the original components. Since the number of significant components is less than the number of initial features, we can limit ourselves to analyzing a small number of principal components, but estimate most of the overall variability. Thus, the original multidimensional space, which in this case has six dimensions, is reduced to a lower-dimensional space. It is essential that the principal components are not correlated, i.e. they are independent of each other. Very often, their independence reflects not only statistical, but also some other important patterns for the researcher.
In the analysis, we used data for all available intact tips with a notch from Kostenok-1, regardless of the type of pen retouching or the absence of its signs. The measurements of 138 items were subjected to statistical processing.
Statistically significant vectors are usually recognized if their eigenvalues exceed one, therefore, their significance exceeds the significance of the average feature. In this case, the first two main components meet this criterion. The first GC describes more than 44% of the total variability (Table 6). The greatest load in it falls on the features describing the main dimensions of the tip: length, width, arc. A somewhat smaller, but still significant, load falls on the length of the pen. There is a positive relationship between the signs. It can be assumed that the first two attributes are determined mainly by the initial parameters of the workpiece. They are closely related to the degree of convexity of the tip edge (arc). In our opinion, this relationship is not accidental and is important for understanding the process of product formation.
The second GC is significantly inferior to the first one in terms of the described variability: it accounts for 27 % of the total variance. In this case, two features related to the morphology of the pen are of the greatest importance : the length of the pen and its angle, with a negative relationship between them. As noted above, both indicators are most sensitive to the process of working out the tool.
According to general technological concepts, the first two GCS correspond to different stages of the" life " of the tips. The first GC reflects the process of manufacturing the tip, the second-the process of using it. Recall that all GCS are statistically independent directions of variability, i.e. their cross-correlation is zero. This purely technical feature of the principal component method in this case fully corresponds to the essential side of the above processes. With this in mind, let us return to the observed relationship between the main tip dimensions and the degree of convexity of its edge. It is the orthogonality of the first two vectors that indicates that the tip arc was formed already in the process of its manufacture and was a reference initial feature. Possible further changes in the shape of the tool during its use did not lead to statistically significant changes in the values of the trait. Probably, this value (arc) was constant for this shape; it was sought and constantly maintained during the use of the tip.
Table 6. Correlation coefficients between the initial features and the first three principal components
Sign
Main components
I
II
Length
0,887
-0,164
Width
0,866
0,025
Pen length
0,618
-0,725
Arc
0,765
0,381
Corner
0,143
0,959
Oval
0,335
0,018
Proper number
2,64
1,620
% of total variability
0,439
0,270
page 25
Figure 3. Location of arrowheads in the space of the first two SCS and examples of tools located at the poles of the vectors.
4. Zones of location of the main tip fractions in the space of the first two SCS.
A and B are the areas of predominant distribution of the tips of fractions A and B, respectively; T is the area of intersection of the values of tips A and B.
A, B, X and 0 are tips related to the corresponding fractions
Consider the position of objects in the space of the first two SCS (Fig. 3). Recall that positive values of I SCS are characterized by products that have a large length and width, as well as a strong convexity of the edge. Products with the opposite combination of characteristics are characterized by negative values of IGA. The second GC, apparently, differentiates the tips according to the degree of their harmony. Harder-worked products that are characterized by a shorter pen length and larger values of its angle are on the positive pole of this vector, and products that "fell out" at the initial stages of the technological path of the tip are on the negative pole.
Obviously, there are no clear groupings of tips in the vector data space. This indicates the fundamental unity of this category of inventory. At the same time, there is a fairly pronounced zoning in the distribution of tips with different types of retouching (Fig. 4). At the same time, the differences between tips with signs of retouching A and B are highly significant for both groups (Table 7). Based on general ideas, two assumptions can be made in connection with this situation. First , we are dealing with a single category of inventory, and the existing differences are due to different degrees of harmony. Second, there are similar, but still initially different forms of products that have become closer to each other in the process of their use and part-time work.
The first assumption seems to be supported by higher values of the angle and lower values of the pen length of the tips of fraction B, as well as extensive zones of transgression of the II GC values of both main fractions of tips. In this case, the arrowheads from fraction B can be interpreted as the most well-developed weapons in the entire sample. However, this interpretation contradicts the direction of differences between the two groups of tips in terms of the values of I GC, which reflects the overall dimensions of the products. At the same time, the tips of fraction B are superior to the tips of fraction A in all dimensions that characterize the workpiece. This situation does not agree with the first assumption: after all, if fraction B were the final shape of fraction A, the size ratio would be inversely observed. Thus, in terms of the sum of arguments, the second point of view looks more reasonable. The process of manufacturing handpieces, according to the analysis, is as follows. Both types (A and B) had a similar billet, but the tips of group B were on average slightly wider and longer; the convexity of the side opposite the recess and the length of the pen were initially set by the size of the billet in both groups, but the convexity in the opposite group was determined by the size of the billet.-
page 26
See Table 7. Average values of I and II HA of tip fractions A and B and values of the t-criterion
Group of Companies
M A
M B
N A
N B
SD A
SD B
t
d.f.
p
I
-0,24
0,40
53
40
1,05
0,92
-3,09
91
<0,01
II
-0,29
0,66
53
40
0,77
0,99
-5,19
91
<0,001
Table 8. Average values of I and II GC fractions of tips B and 0 and values of the t-criterion
Group of Companies
M B
M0
N B
N0
SD B
SD0
t
d.f.
p
I
0,40
-0,21
40
22
0,92
0,85
2,58
60
<0,05
II
0,66
-0,53
40
22
0,99
0,84
4,78
60
<0,001
Table 9. Average values of I and II HA of tip fractions A and 0 and values of the t-criterion
Group of Companies
M A
M0
N A
N0
SD A
SD0
t
d.f.
p
I
-0,24
-0,21
53
22
1,07
0,85
-0,13
73
NS
II
-0,29
-0,53
53
22
0,78
0,84
1,22
73
NS
the number of cores in group B was significantly larger and corresponded to a larger width of the blank.
During the operation of the tips of both groups, the pen length decreased and its angle increased. This change cannot be attributed to the features of the blank - there is no reliable relationship between these features. As already mentioned, these signs may have changed during the use of the handpiece and its adjustment. The angle became wider for the tips with traces of flat retouching (B), but the tendency to change in both groups was the same. The peculiarity of secondary shaping in group B was manifested in an increase in the convexity of the side with an increase in the angle of the pen. The relationship between these features is due to a certain shift of the upper point of the pen away from the arc with a significant increase in the angle and a decrease in the tip length. The absence of a negative statistical relationship between the values of the arc, pen and tool length in group B can only be explained by the lack of initial standards for the length of the workpiece used, i.e. the initial length of the tip of group B could be different. This did not significantly affect the production and use of the handpiece with traces of flat retouching. Otherwise, these signs are related in group A; here there is a negative relationship between the angle and the length of the tip. In other words, the original length of the gun was more standardized. The above differences in the combination of features may capture a functional feature of the tips of group B-a feature that manifests itself against the background of similar processes of use and adjustment.
Of course, there is a possibility that our sample is incomplete and large items with traces of retouching A are simply not represented in it, but according to our data, a situation similar to the one described above is typical not only for Kostenok-1, but also for other monuments where two forms of arrowheads were found (Avdeevo, Zaraiskaya parking lot).. Without claiming to define the functional use of handpieces, we can make a comment based on the regularities of morphological analysis. In the presented sample, the pen is constantly changing in the direction of increasing its angle and decreasing its length, while maintaining the oval and arc of the entire tip. There was a standard billet, but the shape parameters were constantly changing, which could hardly correspond to the function of a throwing weapon. This variability is especially characteristic for tips with signs of flat ventral retouching. For them, an increase in the angle led to a curved side. There are known samples in which the measure of curvature (arc) exceeded the possibility of forming a pen; the tip received the shape of a segment. The variability of the shape most likely corresponds to another function of the tool-the function of the knife, which was described by S. A. Semenov [1957, pp. 116-118].
It remains for us to discuss the relationship of the two tip shapes discussed above with products that do not have traces of retouching on either the dorsal or ventral surfaces of the pen (fraction 0). The size of this group (22 individuals) allows us to make some general observations. This category of inventory can give a general idea of the variability of the initial shape of any main group of handpieces. If such a connection is not found, we will have to admit that these products are the starting form for artifacts of both main groups. Since, despite the specifics of the tips of fractions A and B described above, the values of the first two GC's for these groups are
page 27
If they overlap, you should statistically evaluate their difference from non-retouched tips. The results of the performed analysis clearly indicate that the inventory of this category in terms of its metric indicators differs sharply from the tips of fraction B (Table. 8), but does not show significant differences from the tips of fraction A (Table. 9) by the values of both GC's. This fact allows us to consider most products without traces of retouching in the latter category. Note that the range of values of I GC for non-retouched products is a large part (approx. 70 %) of the same indicator for tips with traces of retouching A, and the range of values for II GC completely covers the variability of the latter group. Since products without signs of retouching, as a rule, have macro traces of harmony, we can say that they were evaluated by a person as fully usable. This indicates a sufficient degree of plasticity of the requirements imposed on the tips of fraction A. Retouching of the tool in this case was considered not as an end in itself, but only as a means to achieve the required shape of the cutting edge of the product.
Conclusions
Based on the results of the performed analysis, we can propose the following scheme for forming Kostenkov tips:
1) the two main fractions of handpieces identified on the basis of differences in the shape of retouching (A and B) are statistically significantly different in their metric characteristics. At the same time, large values for all indicators are typical for products that have traces of flat retouching on the abdominal side (B).;
2) according to the results of the multidimensional analysis, the shape of the tip initially had a high level of variability; it was determined by the shape and size of the workpiece and was set already at the stage of manufacturing tools. About a third of the signs of variability appeared at the stage of using the tool and correcting it. This process mostly affected two metrics-the angle and length of the pen. The absence of noticeable changes in the arc and oval of the tip indicates the technological importance of these elements; their values were maintained at a certain level during the use of the tool;
3) the large average size of the tips with signs of retouching B with higher harmony rates, compared to similar indicators for products with traces of retouching A, allow us to talk about the initial independence of these fractions. Metric convergence is probably the result of similarities in the technological methods of their adjustment;
4) the fraction of non-retouched tips does not differ in its metric characteristic from the fraction of tips with signs of retouching A and should be considered as a shape close to the original one for this group of products;
5) based on statistically processed morphological features, it is possible to speak with a sufficient degree of confidence about the constant variability of the main elements of the tip - an increase in the angle of the pen, a decrease in its length, an increase in the convexity of the side and the presence of traces of flat retouching on changing areas of the tool. This process of constant variability cannot correspond to the use of the tip as a throwing tool. S. A. Semyonov's suggestion about using the Kostenkov tip with a notch as a knife seems more acceptable to us.
List of literature
Amirkhanov H. A. Vostochny gravett or gravettoid industries of Central and Eastern Europe? // Vostochny gravett / Ed. by Kh. A. Amirkhanov, Moscow: Nauch. mir Publ., 1998, pp. 15-33.
Amirkhanov Kh. A. Zaraiskaya stoyanka, Moscow: Nauch. mir Publ., 2000. - 248 p.
Belyaeva V. I. Kremnevyi inventory Kostenok-1 (Opyt klassifikatsii): Avtoref. dis. ... kand. ist. nauk. - L., 1979. - 19 p.
Gvozdover, M. D., Specific features of the flint inventory of the Avdiivka Paleolithic Site, KSIA, No. 82, pp. 112-11961.
Gvozdover M. D. Flint inventory of the Avdiivka Upper Paleolithic site // Vostochny gravett / Ed. by Kh. A. Amirkhanov, Moscow: Nauch. mir Publ., 1998, pp. 234-277.
Gvozdover M. D., Rogachev A. N. Razvitie verkhnepaleoliticheskoi kul'tury [Development of the Upper Paleolithic culture], in Loess, pereglacial, paleolithic on the territory of Middle and Eastern Europe, ed. by I. P. Gerasimov, Moscow: Publishing House of the USSR Academy of Sciences, 1969, pp. 481-530.
Efimenko P. P. Kostenki-1. - Moscow; L.: Publishing House of the USSR Academy of Sciences 1958. - 483 p.
Lev S. Yu. Stone inventory of the Zaraisk parking lot (typological aspect): Abstract of the dissertation of the Candidate of Historical Sciences, Moscow, 2003, 28 p.
Rogachev A. N., Praslov N. D., Anikovich M. V., Belyaeva V. I., Dmitrieva T. N. Kostenki-1 / Ed. by N. D. Praslov, A. N. Rogachev. Moscow, Nauka Publ., 1982, pp. 42-66.
Semenov S. A. Pervobytnaya tekhnika [Primitive technology], Moscow; L.: Publishing House of the USSR Academy of Sciences, 1957, 240 p. (MIA; N 54).
Trusov A.V. Flint complex of the Zaraisk Paleolithic site // Vostochny gravett / Ed. by Kh. A. Amirkhanov, Moscow: Nauch. mir Publ., 1998, pp. 279-298.
Demars P-Y., Laurent P. Types d'outils lithiques du paleolithique superieur en Europe. - P.: CNRS Editions, 2000. - 178 p.
The article was submitted to the Editorial Board on 26.12.06.
page 28
New publications: |
Popular with readers: |
News from other countries: |
 |
Editorial Contacts |
About · News · For Advertisers |
 Digital Library of Uzbekistan ® All rights reserved.
2020-2025, BIBLIO.UZ is a part of Libmonster, international library network (open map) Keeping the heritage of Uzbekistan |
|
US-Great Britain
Sweden
Serbia
Russia
Belarus
Ukraine
Kazakhstan
Moldova
Tajikistan
Estonia
Russia-2
Belarus-2
