Journal of Human Evolution xxx (2014) 1e8

Contents lists available at ScienceDirect

Journal of Human Evolution

journal homepage: www.elsevier.com/locate/jhevol

Fire at will: The emergence of habitual re use 350,000 years ago

Ron Shimelmitz a, *, Steven L. Kuhn b, Arthur J. Jelinek b, Avraham Ronen a, Amy E. Clark b,
Mina Weinstein-Evron a
a
b

Zinman Institute of Archaeology, University of Haifa, 199 Aba Hushi Avenue, Mount Carmel, Haifa 3498838, Israel
School of Anthropology, University of Arizona, 1009 E. South Campus Drive, Tucson, AZ 85721, USA

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 1 October 2013
Accepted 8 July 2014
Available online xxx

The use of re is central to human survival and to the processes of becoming human. The earliest evidence for hominin use of re dates to more than a million years ago. However, only when re use became
a regular part of human behavioral adaptations could its benets be fully realized and its evolutionary
consequences fully expressed. It remains an open question when the use of re shifted from occasional
and opportunistic to habitual and planned. Understanding the time frame of this 'technological mutation'
will help explain aspects of our anatomical evolution and encephalization over the last million years. It
will also provide an important perspective on hominin dispersals out of Africa and the colonization of
temperate environments, as well as the origins of social developments such as the formation of provisioned base camps. Frequencies of burnt ints from a 16-m-deep sequence of archaeological deposits at
Tabun Cave, Israel, together with data from the broader Levantine archaeological record, demonstrate
that regular or habitual re use developed in the region between 350,000e320,000 years ago. While
hominins may have used re occasionally, perhaps opportunistically, for some million years, we argue
here that it only became a consistent element in behavioral adaptations during the second part of the
Middle Pleistocene.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Habitual re
Lower Paleolithic
Acheulo-Yabrudian
Tabun Cave
Mount Carmel

Introduction
The available archaeological record shows that evidence for the
use of re, whether well-controlled or not, is highly sporadic for
most of the Lower Paleolithic period (2.6 million-250 kya [thousands of years ago]) (e.g., Alperson-Al and Goren-Inbar, 2010;
Berna et al., 2012; Gowlett and Wrangham, 2013) with some evidence for intensication at ca. 300e400 kya (e.g., Rolland, 2004;
de Lumley, 2006; Karkanas et al., 2007; Roebroeks and Villa, 2011).
In contrast, there is abundant evidence for habitual use of re
during the Middle Paleolithic (ca. 250e50 kya) (e.g., Albert et al.,
1999; Meignen et al., 2007; Berna and Goldberg, 2008;
Sandgathe et al., 2011a; Albert et al., 2012; Aldeias et al., 2012;
Goldberg et al., 2012; Mentzer, 2014; Weinstein-Evron et al.,
2012). Wrangham and Carmody (2010), who based their argument
on biological models, as well as Gowlett and Wrangham (2013),
who used a wider perspective including biological and archaeological evidence, assert that the available archaeological record

* Corresponding author.
E-mail
addresses:
rshimelm@campus.haifa.ac.il,
(R. Shimelmitz).

ronishim@gmail.com

fails to reveal a clear point in time when re became a regular

component in hominin behavioral repertoire. Currently there is a
diversity of opinion on the topic. Some argue for the onset of
habitual re use at the beginning of the Middle Pleistocene or
earlier (>700 kya) (e.g., Alperson-Al, 2008; Alperson-Al and
Goren-Inbar, 2010; Wrangham and Carmody, 2010; Gowlett and
Wrangham, 2013; Twomey, 2013), whereas others believe that it
developed only around 300e400 kya (e.g., Rolland, 2004; de
Lumley, 2006; Karkanas et al., 2007; Roebroeks and Villa, 2011;
Shahack-Gross et al., 2014). A few scholars argue for a much
later date for the ability to regularly kindle re, sometime after
250 kya (Sandgathe et al., 2011b).
Identifying when habitual re use emerged is of major importance for understanding human evolutionary history. Whereas re
offered many potential advantages to hominins (e.g. Pyne, 2001; de
Lumley, 2006; Stiner et al., 2011), its full impact on biological and
behavioral evolution would have been felt only when re became a
routine component of hominin lifeways (Rolland, 2004). The exact
time frame in which re became a regular part of human behavior
is thus central to our reconstruction of several key features of
evolutionary history, including changes in anatomy (e.g.,
Wrangham, 2009; Wrangham and Carmody, 2010), the dispersal of
hominins into temperate regions (e.g., Gowlett, 2006; Alperson-

http://dx.doi.org/10.1016/j.jhevol.2014.07.005
0047-2484/ 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

Figure 1. The 16-m long sequence of archaeological layers examined from Tabun Cave, Israel. The illustration (A) marks the location of three excavation expeditions at the site:
D.A.E. Garrod (1929e1934), A. Jelinek (1967e1971), and A. Ronen (1975e2003). Garrod's and Jelinek's sections face north while Ronen's lower section faces east. The part of the
section where the transition into habitual use of re occurred is marked with details (B) (light gray marks the layers with dense burnt items).

Al, 2008; Alperson-Al and Goren-Inbar, 2010; Gowlett and

Wrangham, 2013), as well as the intensied social interactions
within base camps (e.g., Rolland, 2004; Stiner et al., 2011).
Fire would have provided many advantages to hominins,
including cooked food, warmth, light and protection from predators (e.g., Rolland, 2004; de Lumley, 2006; Wobber et al., 2008;
Wrangham, 2009; Wrangham and Carmody, 2010; Gowlett and
Wrangham, 2013). The many advantages of re lead us to two
suppositions. First, we assume that re became a regular, even
permanent part of their adaptations once hominins had solved the
technical challenge of kindling and maintaining it. This corresponds
with the prediction that habitual re should appear in the paleoanthropological record as a punctuated event (Rolland, 2004).
Second, as a behavioral feature deeply embedded in human culture
(Rolland, 2004; Sandgathe et al., 2011a) the habitual use of re
should be manifest across many broadly contemporaneous sites

that share other cultural and technological features (Sandgathe

et al., 2011b). Following these studies we refer to habitual re as
an action embedded in human behavior and culture. Archaeologically, we expect habitual use of re to be manifest repetitively at a
cluster of sites belonging to the same time period or cultural
complex. The search for the archaeological manifestation of this
behavior have focused on a single location (e.g., de Lumley, 2006) or
have taken a very broad view, incorporating the most reliable case
studies from particular continents (e.g., Rolland, 2004; Roebroeks
and Villa, 2011). In this study we adopt a middle-ground perspective. We begin by concentrating on one very long, semi-continuous
sequence (Tabun Cave) but then expand out to incorporate evidence from semi-contemporaneous sequences from across the
Levant. In this way we are able to document changes in use of re
within a regional population rather than a single site or a widely
scattered set of sites.

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

Because of its exceptionally long stratigraphic sequence,

covering the Lower Paleolithic through the late Middle Paleolithic,
Tabun Cave, Mount Carmel (Garrod and Bate, 1937; Jelinek et al.,
1973; Jelinek, 1982; Ronen et al., 2011), constitutes an ideal locale
for tracking changes in the frequency and regularity of re use
among Pleistocene hominins. The fact that the site is a cave also
reduces the likelihood of res within the archaeological sediments
having been caused by natural agents (Roebroeks and Villa, 2011).
The rich archaeological record of the Levant further enables us to
check for corroborative evidence at multiple sites. Combining these
two sets of data enables us to more clearly identify the point in time
at which re use became a regular and widespread part of human
behavior, and to overcome the difculties former studies encountered in trying to track consistency in re use across time and space.
Tabun Cave was rst excavated in 1929e1934 by D.A.E Garrod,
who exposed a 25 m deep sequence of archaeological layers
spanning the Lower to the Middle Paleolithic (Garrod and Bate,
1937). Investigations were resumed in 1967e1971 by A. Jelinek
(Jelinek et al., 1973; Jelinek, 1982), who re-excavated a 10 m deep
sequence at the middle of the prole left by Garrod's excavation,
and later by A. Ronen (Ronen et al., 2011), who excavated an
additional 6 m below Jelinek's trench, reaching bedrock. In this
study we combine results from Jelinek's and Ronen's excavations,
together forming a sequence of 100 archaeological layers spanning
a 16 m depth of sediments and representing ca. 500 ky (Mercier
et al., 2000; Mercier and Valladas, 2003) (Fig. 1). This includes
layers J1-J85 and J90 from Jelinek's excavation (Jelinek et al., 1973;
Jelinek, 1982) and layers R62-R75 [original layer record; e.g.,
Shifroni and Ronen, 2000] from Ronen's excavation (Ronen et al.,
2011). In order to avoid confusion, layers from Jelinek's and
Ronen's excavations are labeled with the letters J and R, respectively. The examined sequence includes: (a) Layers R75-R64
(Shifroni and Ronen, 2000) from the Lower Paleolithic Acheulian
(e.g., Bar-Yosef, 1994). These are undated but located below Unit
XIV, which is dated to 415 27 kya (Mercier et al., 2000); (b) Layers
R63-R62 and J90-J70 (Jelinek, 1982, 1990; Shifroni and Ronen,
2000; Shimelmitz et al., 2014) from the Acheulo-Yabrudian complex of the late Lower Paleolithic (e.g., Copeland, 2000), which are
dated from 415 27 to 264 28 kya (Mercier and Valladas, 2003);
(c) Layers J69-J1 (Jelinek et al., 1973; Jelinek, 1982; Shimelmitz and
Kuhn, 2013) from the Middle Paleolithic Levantine Mousterian (e.g.,
Shea, 2003), dated to between 256 26 and 165 16 kya (Mercier
and Valladas, 2003).
Previous studies in Tabun identied the use of re through eld
observations in Jelinek's Unit's I-II (Jelinek et al., 1973) and possibly
Garrod's Layer E (Garrod and Bate, 1937). Traces of re were also
detected through geoarchaeological analyses in Mousterian Units
I-II (Albert et al., 1999) and in the Acheulo-Yabrudian layers toward
the middle-lower part of the sequence (Tsatskin, 2000). The successful thermoluminescence dating conducted on burnt ints
retrieved from Units IeXIV (Mercier et al., 2000; Mercier and
Valladas, 2003) is another indication of the use of re at the site
during the late Lower Paleolithic and the Middle Paleolithic.
Materials and methods
We examined the use of re at Tabun based on the presence of
burnt int items (ints with visible signs of heat alteration)
(Alperson-Al, 2008; Alperson-Al and Goren-Inbar, 2010; Aldeias
et al., 2012). Results of geoarchaeological research (Albert et al.,
1999; Tsatskin, 2000) on Tabun are also incorporated into our
study. Globally there is no evidence of intentional heat-treatment
of int during the Middle and Lower Paleolithic in Eurasia
(although it does appear in Africa in the Middle Stone Age at ca.
160 kya; Brown et al., 2009). Thus, we can assume that ints were

burnt accidentally, through kindling res on top of previously

abandoned artifacts or via discarding debris into hearths. Studies of
the distribution of burnt ints further demonstrate that burnt artifacts are found not only within hearths but also scattered
throughout the general area of excavations, a result of processes
such as cleaning out of replaces, trampling and earth moving (e.g.,
Sergant et al., 2006; Alperson-Al, 2008; Alperson-Al and GorenInbar, 2010; Aldeias et al., 2012). Consequently, when re was used
at the site, burnt material is expected to appear in a broad area. As
such, the frequency of burnt ints should be a suitable proxy for the
frequency of res within the cave. Finally, because stone is so durable, burnt ints are not expected to suffer from the same degree
of post-depositional alteration and destruction as are hearth features, heated sediments, charcoal and ash or even burnt bones.
Our analysis refers only to int items larger than 2.5 cm.
Although examining smaller items might be fruitful as well (e.g.,
Alperson-Al, 2008; Alperson-Al and Goren-Inbar, 2010), the
cemented sediments in some layers prevented excavators from
collecting smaller debris. Additionally, focusing on larger objects
avoids some potential biasing effects of thermal fracture, whereby a
single burnt piece could be transformed into hundreds of tiny burnt
fragments.
Only layers with samples of 300 or more items were incorporated into our study. The layers from Jelinek's Units III-VIII were
excluded as well because they have been reworked by erosion,
especially Units III-VI (Jelinek et al., 1973; Jelinek, 1982). From
Ronen's excavation we included only the layers directly underlying
Jelinek's excavations so as to minimize possible problems in
correlating layers.
Burnt artifacts were divided into two categories: extremely and
moderately burnt. The former bears characteristics typical of
exposure to heat levels
300  C, including changes in color and
texture, pot-lid fractures and crazing (Sergant et al., 2006). Thermoluminescence studies of ints from Tabun cave support their
alteration by re (Mercier et al., 2000; Mercier and Valladas, 2003).
Artifacts belonging to the moderately burnt' group bear some of
the same signs of heating but they are fewer and less well
expressed. To minimize ambiguities in the search for combustion
features through the spatial analysis (using ArcGIS 10) we focused
on the extreme burnt items.
Results
Of the 100 layers, we examined a total of the 27 larger assemblages (>300 items) from Tabun, incorporating a sample of 43,405
items (Fig. 2), constituting ca. 80% of the nds retrieved from this
locale. The frequencies of 'extremely' and 'moderately' burnt ints
show a strong positive correlation across the layers studied (Pearson's r 0.836; p < 0.001), conrming that the 'moderately' heated
items primarily resulted from exposure to heat. Accordingly, we
incorporated the items in both categories into our analysis.
Marked changes in frequencies of burnt items over the long
stratigraphic sequence at Tabun indicate clear uctuations in the
occurrence of re within the cave. Within the lower layers of the
site, layers R66 to J82, attributed to the Lower Paleolithic Acheulean
and early Acheulo-Yabrudian complex, burnt int items are practically absent, with 'extremely' burnt int items constituting just
0e0.7% of the total assemblage. Adding the 'moderately' burnt
items increases the total of burnt int items up to 1.3% only. Layers
R75-R67, just above bedrock, are not included in this study since
the assemblages are very small. Nevertheless, it is noteworthy that
only isolated burnt ints were found in these layers. The AcheuloYabrudian Layer J81 (in the middle of the sequence, 8 m below
datum and 8 m above bedrock) marks the rst real departure from
the trend seen in the earlier levels. Burnt ints constitute ca. 10% of

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

60
50
40
30
20

Lower Paleolithic, Acheulean (Tabun F)

(Tabun C)

70

moderate heat
extreme heat

Late Lower Paleolithic, Acheulo - Yabrudian

(Tabun E)

R64

R63

J90

J84

J82

J83

J81

J80

J77

J79

J76

J75

J73

J74

J72

J68

J71

J66

J25

J63

J23

J22

J20

J19

R62

10

2043

1317

990

1340

5793

1951

4592

2582

n=

1318

Undated

2078

41527

2317

30227

445

32431
3131

26428
4527

26722

1524

25626

1097

16516

1075

Kya

436

No designation
to units

325

XIV

379

XIII

321

XII

859

XI

560

618

IX

371

878

Unit

R66

80

Early Middle Paleolithic (Tabun D)

Middle
Middle
Paleolithic

R65

90

466

Figure 2. Fluctuations in frequencies of ints (
2.5 cm) with heat alteration throughout the sequence of Tabun Cave demonstrating the changing intensity of re use. The division
into units and layers as well as their mean dates (Mercier et al., 2000; Mercier and Valladas, 2003) are presented. Since the date of Unit XIII is an anomaly we prefer to use the single
date of Bed 83 instead. A sharp transition appears in Layer J81, roughly dated to 357e324 kya. Below this layer burnt ints are scarce or absent but above it they are a constant
feature of the assemblages.

the assemblage in layer J81. Because there is an unconformity at the

top of layer J81 it is possible that the burning actually occurred
during the deposition of layer J80, but this makes little difference in
terms of chronology. Moving upwards through the sequence, proportions of burnt ints uctuate within a relatively limited range
(8.4e23.1%) in the overlying layers associated with the AcheuloYabrudian complex (J80-J71). Although the individual layers are
relatively thick (ca. 30e40 cm) and most likely represent a series of
occupations rather than a single event, in some of these layers
patterns in the distribution of 'extremely' burnt ints can still be
detected (Fig. 3).
Since burnt ints crack and shatter into numerous pieces, we
also provide a different perspective that includes the presence of
burnt material (extreme and moderate) solely among whole and
proximal blanks (including shaped items) (Fig. 4). This was conducted on a cluster of several sub-beds that are currently under
technological analysis and not on the entire set of assemblages.
While the results show slightly different values than in Fig. 2, the
general trend is the same.
The results of geoarchaeological studies at Tabun (Tsatskin,
2000) are also consistent with the observed differences in
burning frequencies across the stratigraphy. A variety of burnt
material including tiny int chips and bone fragments as well as
heated soil aggregates was observed within the micromorphological thin-sections from layers attributed to Units XII-XIII but no such
traces of re were found in Unit XIV (Layer J90). We note that
changing frequencies of burnt ints at Tabun are not simply a
consequence of uctuating densities of ints within the deposits. In
fact, within Units XI-XIV, there is a weak but negative relationship
between the frequencies of burnt ints and overall nd density
(r 0.397, df 93, p < 0.001).
Based on existing thermoluminescence dates, we can conclude
that habitual re use at Tabun began sometime before 324 31 kya,
the mean date of Unit XII (Mercier and Valladas, 2003). The upper

part of Unit XIII suffered from a certain degree of erosion before

Unit XII began forming, which may explain the problematic
younger mean date (302 27 kya). Of the four thermoluminescence dates obtained from Unit XIII, three (T63, T64, T68) were
from layer J81, which was most affected by this erosion and
exposure. The remaining sample (T67) dated to 357 33 kya came
from a deeper layer (J83) and puts a lower limit on the documented
shift in pyrotechnology. Thus, we date the emergence of habitual
use of re at Tabun to somewhere between 357 and 324 kya.
The trends in burning frequencies at Tabun are not monotonic.
The early Middle Paleolithic deposits, Layers J68eJ63, actually
contain a smaller proportion of burnt ints than some earlier layers,
although the regular use of re is clearly evident. There is a dramatic rise in the percentage of burnt ints in Layers J25eJ19 of the
mid-Middle Paleolithic. The higher frequencies of burnt pieces in
these layers accord with the presence of clearly visible black and
white hearth layers with traces of burnt wood in the sediments
(Albert et al., 1999).
Discussion
The scarce evidence of re at the lowest layers of Tabun Cave
attributed to the Acheulean and early Acheulo-Yabrudian ts the
general pattern within Eurasia, where similarly early re use was
reported from a few isolated cases (Roebroeks and Villa, 2011). In
order to examine whether the trend in re use within the Tabun
sequence reects a regional rather than a local pattern we examined published evidence from other sites and layers associated
with the Acheulo-Yabrudian complex. The advantages of examining the Acheulo-Yabrudian complex is that it is conned to a
relatively limited geographic area, the eastern Mediterranean
Levant, and is restricted to the later part of the Middle Pleistocene,
ranging between 415 and 250/220 kya (Mercier et al., 2000;
Mercier and Valladas, 2003; Gopher et al., 2010; Valladas et al.,

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

Figure 3. Density maps of heavily burnt ints in Beds J79 (aec) and J80A (dee), Unit XII, dated to 324 kya. Bed J80A constitutes the upper part of layer J80 (n 1470). Top row (a, d)
contains density maps of all artifacts; middle row (b, e) contains density maps of heavily burnt pieces; bottom row show the result of dividing the heavily burnt raster map by the
raster of all artifacts, analogous to the local proportion of burnt pieces. The maps show areas with particularly high frequencies of burnt artifacts (circled in red) in both layers, which
may be indicative of traces of combustion features or rake out from nearby hearths. (For interpretation of the references to colour in this gure legend, the reader is referred to the
web version of this article.)

2013). Almost all Acheulo-Yabrudian assemblages come from cave

sites, hence the depositional contexts are also relatively similar to
Tabun.
Qesem Cave, which spans the period of ca. 400e220 kya
(Gopher et al., 2010; Mercier et al., 2013) is one of the best comparators for the middle part of the Tabun sequence. There is a
marked difference between the upper and lower parts of the Qesem
sequence. Although burnt ints and bones were present in the
lower layers, geoarchaeological analysis revealed only few combustion products in sediments; this is in marked contrast to the
upper part of the sequence, where the sediments were packed with
ash and other combustion products (Karkanas et al., 2007). A
marked increase in the frequency of burnt bones within the lower
sequence at Qesem cave (Stiner et al., 2011) seems to parallel the
observations on burnt ints from Tabun. The newly reported hearth
from Qesem, dated to ca. 300 kya is further evidence of this transition in the use of re at the cave (Shahack-Gross et al., 2014).

Yabrud I, Syria may also document a similar trend, although the

data are more coarse-grained. Yabrud I contains a long sequence of
Lower and Middle Paleolithic levels. Farrand (1965: 39) examined
Rust's (1950) section and mentioned six discontinuous hearth
levels in elevation 2.5e3.3 m and another very prominent hearth
layer, almost 10 cm thick at elevation 4.5 m. This was also
conrmed by de Heinzelin (1966). These elevations correlate to
the later part of the Acheulo-Yabrudian layers at the site (Layers
11e18; Rust, 1950). During the excavations of Solecki and Solecki
several additional hearths were identied within the AcheuloYabrudian layers, some of them were associated with sets of
stones which appeared to have encircled (the) hearths
(Solecki, 1970: 204). These reports mention no evidence of re in
the lower part of the sequence of Yabrud I (Layers 19e25; Rust,
1950).
Evidence from other Acheulo-Yabrudian sites with shorter
stratigraphic sequences further conrms that use of re was a

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

R66

R65

R64

R63

J90C1+J90H1

all items

J83B1

J82BS

J79A1+J79AIA+

J76I2B

J72S

J75I1+J75I2+J75S

whole and proximal

18
16
14
12
10
8
6
4
2
0

Figure 4. Percentages of heat alteration among whole items and proximal fragments
(including shaped/retouched items) from Layers R63-R66 and from specic sub-beds
from layers J72, J75, J76, J79, J82, J83 and J90. All items n: J72S 479;
J75I1J75I2J75S 2059; J76I2B 825; J79A1J79AIA J79A2 J79A3 832;
J82BS 514; J83B1 1011; J90C1J90H1 555; R63 990; R64 1317; R65 2043;
R66 466. Whole and proximal items n: J72S 282; J75I1J75I2J75S 1348;
J76I2B 332; J79A1J79AIA J79A2 J79A3 488; J82BS 350; J83B1 678;
J90C1J90H1 332; R63 662; R64 907; R65 1322; R66 278.

generalized behavioral pattern during that period. Geoarchaeological studies identied traces of re in all of the cave sites
examined that preserve Acheulo-Yabrudian layers (Table 1). The
only sites where re was not recognized are open-air springmound locations in Syria (Le Tensorer et al., 2007). Although the
cited studies used different methods, the composite picture is that
re was an integral part of hominin activities in cave sites of the
later Acheulo-Yabrudian complex across the eastern Mediterranean
region. It is also important that Qesem, Tabun and possibly Yabrud
all show that a more regular use of re emerged within the
Acheulo-Yabrudian complex rather than at its beginning. Thus, it is
not associated with a major change in material culture or the
appearance of new hominin population. In addition, while traces of
re might be expected to be more common in cave sites as opposed
to open-air sites, the Levantine evidence does not simply correlate
with more frequent use of caves. Whereas most of the earlier
Acheulean sites are open-air localities (e.g., Bar-Yosef, 1994),

increases in the regularity of burning post-date the earliest

Acheulo-Yabrudian cave occupations at Tabun, Qesem and Yabrud.
By roughly 200 kya re was commonly used wherever hominins
were able to survive (Rolland, 2004; Roebroeks and Villa, 2011;
Mentzer, 2014). Evidence pertaining to the frequency of re varies throughout the Middle Paleolithic of the Levant. While some
sites show a slight decline in the amount or intensity of burning
during the early Middle Paleolithic, as in Tabun, others show
distinct hearths and pronounced burning activities as demonstrated by the frequent burnt ints and other materials (e.g.,
Weinstein-Evron et al., 2012). The intensive use of re in the assemblages of Tabun Unit I, reected in the exceptionally large
quantities of burnt ints may be connected to another transformation in the use and management of re, but it may also simply
reect a more continuous occupation of this particular cave.
Conclusions
The ubiquity of evidence from Acheulo-Yabrudian sites in the
Levant makes it clear that re became a regular part of hominin
behavior during the second half of the Middle Pleistocene. Given
the available dates, our best estimate for the onset of regular re
use at Tabun is between 357 and 324 kya. These results are in
agreement with the recent review of early evidence of re from
Europe and around the globe (Rolland, 2004; Roebroeks and Villa,
2011), which concludes that the earliest sporadic traces of re in
Europe appear during MIS 11-10, with an increase in the frequency
of evidence in MIS 9. However, the ndings from the Levant provide
a density of coverage in time and space that is currently unavailable
elsewhere. Consistent evidence for re is found not just in the
Tabun sequence but at every Acheulo-Yabrudian cave site where
good information is available. The near-absence of burnt ints in
the lower 8 m of the sequence at Tabun (composed of 19 layers) also
indicates that the scarcity of re evidence before 350 kya is not just
a matter of spotty preservation, or cave sites versus open air-sites
(e.g., Gowlett and Wrangham, 2013). Rather, the negative evidence from the early layers is genuine, and there is a signicant and
permanent increase in the frequency of evidence for burning between 357 and 324 kya. This is the period in which the technological mutation underlying habitual use of re emerged. We
suggest that the changes in burning frequency at Tabun and Qesem
not only signal the point in time where the use of re became
habitual, but also indicate that humans had mastered the art of
kindling re. Unfortunately there are no means currently available
to directly determine how ancient res were started, so the latter
remains simply a hypothesis for the time being.

Table 1
Evidence of re in sites of the Acheulo-Yabrudian complex, generally dated to 415e250/220 kya.
Site

Reference
a

Jamal
Hayonim
Yabrud I
Misliya
Bezez
Abri Zumoffen
Masloukh

Mercier et al., 2000; Tsatskin, 2000; Mercier and Valladas, 2003;

Rink et al., 2004
Karkanas et al., 2007; Gopher et al., 2010; Stiner et al., 2011;
Mercier et al., 2013
Weinstein-Evron and Tsatskin, 1994; Weinstein-Evron et al., 1999
Mercier et al., 2007; Berna and Goldberg, 2008
Farrand, 1965; de Heinzelin, 1966, Solecki, 1970; Porat et al., 2002
Valladas et al. 2013
Kirkbride, 1983
Copeland, 1983
Skinner, 1970

Hummal
Nadaouiyeh

Le Tensorer et al., 2007

Le Tensorer et al., 2007

Tabun

Qesem

Date kya

Fire

Evidence type

415-250 TL 387 ESR/US

Micromorphology; burnt int

217-308(421) TL 225-388 ESR/US

194/220-320/420US
>220 UeTh
>230 TL

220 ESR
250-290 TL

Micromorphology; burnt bones;

burnt int
Micromorphology
Micromorphology
Geoarchaeology; hearths
Burnt int
Field observation: hearth
Field observation: hearth
Field observation:
burnt bones and charcoal
Micromorphology
Micromorphology




The only two sites without evidence of re are open-air locations, not caves.
a
Dates of Tabun refer only to those obtained from Jelinek's and Ronen's excavations.

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

Initial dates for any historical phenomenon are minimum estimates and new discoveries almost inevitably push the ages of
rsts back in time. The archaeological signal is much stronger
when behaviors become widespread and habitual. While the
earliest evidence of re associated with hominin activities is much
older, the data presented here indicate that re became a regular
and constant part of hominin behavioral adaptations in Eurasia
only after 350 kya. The benets of re for processing food, altering
raw materials or enhancing social interactions would be fully
realized only when use of re shifted from opportunistic and occasional to habitual and regular. Regular use of re changed hominin existence and inuenced the direction of evolution in our
lineage in a diversity of ways (Rolland, 2004; Wrangham and
Carmody, 2010; Stiner et al., 2011). To the extent that humans'
physical, cognitive and social evolution was affected by the emergence of habitual re use, we should be able to trace evidence for
such impacts most clearly in the hominins that inhabited the
landscape over the last 350 kya.
Acknowledgments
The study was conducted as part of a post doctoral study (R.S.) at
the Zinman Institute of Archaeology, University of Haifa. The new
examination of material from Tabun Cave is supported by the Irene
Levi Sala Care Archaeological Foundation. We wish to thank two
anonymous reviewers. Thanks is also due to Anat Regev for drawing
Figures 1-2.
References
Albert, R.M., Lavi, O., Estroff, L., Weiner, S., 1999. Mode of occupation of Tabun Cave,
Mt Carmel, Israel during the Mousterian period: a study of the sediments and
phytoliths. J. Archaeol. Sci. 26, 1249e1260.
Albert, R.M., Berna, F., Goldberg, P., 2012. Insights on Neanderthal re use at Kebara
Cave (Israel) through high resolution study of prehistoric combustion features:
evidence from phytoliths and thin sections. Quatern. Int. 247, 278e293.
Aldeias, V., Goldberg, P., Sandgathe, D., Berna, F., Dibble, H.L., McPherron, S.P.,
Turq, A., Rezek, Z., 2012. Evidence for Neandertal use of fire at Roc de Marsal
(France). J. Archaeol. Sci. 39, 2414e2423.
Alperson-Al, N., 2008. Continual re-making by hominins at Gesher Benot Ya'aqov,
Israel. Quatern. Sci. Rev. 27, 1733e1739.
Alperson-Al, N., Goren-Inbar, N., 2010. The Acheulian Site of Gesher Benot Ya'aquv
Volume II: Ancient Flame and Controlled Use of Fire. Springer, New York.
Bar-Yosef, O., 1994. The Lower Paleolithic of the Near East. J. World Prehist. 8/3,
211e265.
Berna, F., Goldberg, P., 2008. Assessing Paleolithic pyrotechnology and associated
hominin behavior in Israel. Israel J. Earth Sci. 56, 107e121.
Berna, F., Goldberg, P., Kolska Horwitz, L., Brink, J., Holt, S., Bamford, M., Chazan, M.,
2012. Microstratigraphic evidence of in situ re in the Acheulean strata of
Wonderwerk Cave, Northern Cape province, South Africa. Proc. Natl. Acad. Sci.
109, E1215eE1220.
Brown, K.S., Marean, C.W., Herries, A.I.R., Jacobs, Z., Tribolo, C., Braun, D.,
Roberts, D.L., Meyer, M.C., Bernatchez, J., 2009. Fire as an engineering tool of
early modern humans. Science 325, 859e862.
Copeland, L., 1983. Abri Zumoffen, 1958, in Retrospect and the Extension, 1963. In:
Roe, D. (Ed.), Adlun in the Stone Age. The Excavations of D.A.E. Garrod in
Lebanon 1958e1963. British Archaeological Reports International Series 159.
Archaeopress, Oxford, pp. 75e88.
Copeland, L., 2000. Yabrudian and related industries: the state of research in 1996.
In: Ronen, A., Weinstein-Evron, M. (Eds.), Toward Modern Humans: Yabrudian
and Micoquian, 400e50 k-years ago. British Archaeological Reports International Series 850. Archaeopress, Oxford, pp. 97e117.
de Heinzelin, J., 1966. Revision du Site de Yabroud. Annales Archeologique Arabes
Syriennes XVI/2, 157e165.
de Lumley, H., 2006. Il y a 400 000 ans : la domestication du feu, un formidable
moteur d'hominisation. C. R. Palevol 5, 149e154.
Farrand, W.R., 1965. Geology, climate and chronology of Yabrud rockshelter I. Les
Annales Archeologiques de Syrie XV, 35e50.
Garrod, D.A.E., Bate, D.M.A., 1937. The Stone Age of Mount Carmel, I. Clarendon
Press, Oxford.
Goldberg, P., Dibble, H., Berna, F., Sandgathe, D., McPherron, S.J.P., Turq, A., 2012.
New evidence on Neandertal use of re: Examples from Roc de Marsal and Pech

IV. Quatern. Int. 247, 325e340.
de l'Aze
Gopher, A., Ayalon, A., Bar-Matthews, M., Barkai, R., Frumkin, A., Karkanas, P.,
Shahack-Gross, R., 2010. The chronology of the late Lower Paleolithic in the

Levant: U series dates of speleothems from Middle Pleistocene Qesem cave,

Israel. Quatern. Geochronol. 5, 644e656.
Gowlett, J.A.J., 2006. The early settlement of northern Europe: re history in the
context of climate change and the social brain. C.R. Palevol 5, 299e310.
Gowlett, J.A.J., Wrangham, R.W., 2013. Earliest re in Africa: towards the convergence of archaeological evidence and the cooking hypothesis. Archaeological
Research in Africa 48/1, 5e30.
Jelinek, A.J., 1982. The Middle Paleolithic in the southern Levant, with comments on
the appearance of modern Homo sapiens. In: Ronen, A. (Ed.), The Transition
from Lower to Middle Paleolithic and the Origin of Modern Man. British
Archaeological Reports International Series 151. Archaeopress, Oxford,
pp. 57e101.
Jelinek, A.J., 1990. The Amudian in the context of the Mugharan Tradition at the
Tabun Cave (Mount Carmel), Israel. In: Mellars, P. (Ed.), The Emergence of
Modern Humans. Cornell University Press, Ithica, pp. 81e90.
Jelinek, A.J., Farrand, W.R., Haas, G., Horowitz, A., Goldberg, P., 1973. New excavation
at the Tabun Cave, Mount Carmel, Israel, 1967e1972: a preliminary report.

orient 1, 151e183.
Pale
Karkanas, P., Shahack-Gross, R., Ayalon, A., Bar-Matthews, M., Barkai, R., Fromkin, A.,
Gopher, A., Stiner, M., 2007. Evidence of habitual use of re at the end of the
Lower Paleolithic: site-formation processes at Qesem Cave. Israel. J. Hum. Evol.
53, 197e212.
Kirkbride, D., 1983. The Soundings at the Mugharat El-Bezez. In: Roe, D. (Ed.), Adlun
in the Stone Age. The Excavations of D.A.E. Garrod in Lebanon 1958e1963.
British Archaeological Reports International Series 159. Archaeopress, Oxford,
pp. 23e52.
Le Tensorer, J.-M., Jagher, R., Rentzel, P., Hauck, T., Ismail-Meyer, K., Pmpin, C.,
Wojtczak, D., 2007. Long-term formation processes at the natural springs
Nadaouiyeh and Hummal in the El-Kowm Oasis, central Syria. Geoarchaeology
22/6, 621e639.
Meignen, L., Goldberg, P., Bar-Yosef, O., 2007. The hearths at Kebara Cave and their
role in site formation processes. In: Bar-Yosef, O., Meignen, L. (Eds.), Kebara
Cave, Mt. Carmel, Israel. Peabody Museum of Archaeology and Ethnology
Harvard University, Cambridge, pp. 91e122.
Mentzer, S.M., 2014. Microarchaeological approaches to the identication and
interpretation of combustion features in prehistoric archaeological sites.
J. Archaeol. Method Th 21, 616e668.
Mercier, N., Valladas, H., 2003. Reassessment of TL age estimates of burnt ints from
the Paleolithic site of Tabun Cave. Israel. J. Hum. Evol. 45, 401e409.
Mercier, N., Valladas, H., Froget, L., Joron, J.-L., Ronen, A., 2000. Datation par ther
olithique de Tabun (Mont Carmel,
moluminescence de la base du gisement pale
l). C.R. Acad. Sci. Paris 330, 731e738.
Israe
Mercier, N., Valladas, H., Froget, L., Joron, J.-L., Reyes, J.-L., Weiner, S., Goldberg, P.,
Meignen, L., Bar-Yosef, O., Belfer-Cohen, A., Chech, M., Kuhn, S.L., Stiner, M.C.,
Tillier, A.-M., Arensburg, B., Vandermeersch, B., 2007. Hayonim Cave: a TL-based
chronology for the Levantine Mousterian sequence. J. Archaeol. Sci. 34/7,
1064e1077.
res, C., Shao, Q., Gopher, A., Barkai, R., Bahain, J.-J.,
Mercier, N., Valladas, H., Falgue
Vialettes, L., Joron, J.-L., Reyss, J.-L., 2013. New datings of Amudian layers at
Qesem Cave (Israel): results of TL applied to burnt ints and ESR/U-series to
teeth. J. Archaeol. Sci. 40, 3011e3020.
Porat, N., Chazan, M., Schwarcz, H., Horwitz, L.K., 2002. Timing the Lower to
Middle Paleolithic boundary: New dates from the Levant. J. Hum. Evol. 43,
107e122.
Pyne, S.J., 2001. Fire: A Brief History. University of Washington, Seattle.
Rink, W.J., Schwartz, H.P., Ronen, A., Tsatskin, A., 2004. Conrmation of a near 400
ka age for the Yabrudian industry at Tabun Cave. Israel. J. Archaeol. Sci. 31,
15e20.
Roebroeks, W., Villa, P., 2011. On the earliest evidence for habitual use of re in
Europe. Proc. Natl. Acad. Sci. 108, 5209e5214.
Rolland, N., 2004. Was the emergence of home bases and domestic re a punctuated event? A review from the Middle Pleistocene record in Eurasia. Asian
Perspective 43, 248e280.
Ronen, A., Gisis, I., Tchernikov, I., 2011. The Mugharan Tradition reconsidered. In: Le
Tensorer, J.-M., Jagher, R., Otte, M. (Eds.), The Lower and Middle Paleolithic in
the Middle East and Neighboring Regions, 126. ERAUL, pp. 59e66.
Rust, A., 1950. Die Hohlenfunde von Jabrud (Syrien). Wachholtz, Neumunster.
Sandgathe, D.M., Dibble, H.L., Goldberg, P., McPherron, S.P., 2011a. On the role of re

IV
in Neandertal adaptations in Western Europe: evidence from Pech de l'Aze
and Roc de Marsal, France. PaleoAnthropology 2011, 216e242.
Sandgathe, D.M., Dibble, H.L., Goldberg, P., McPherron, S.P., Turqh, A., Nivend, L.,
Hodgkinse, J., 2011b. Timing of the appearance of habitual re use. Proc. Natl.
Acad. Sci. 108, E298.

, P., Perdaen, Y., 2006. The invisible hearths: a contribution to
Sergant, J., Crombe
the discernment of Mesolithic non-structured surface hearths. J. Archaeol. Sci.
33, 999e1007.
Shahack-Gross, R., Berna, F., Karkanas, P., Lemorini, C., Gopher, A., Barkai, R., 2014.
Evidence for the repeated use of a central hearth at Middle Pleistocene (300 ky
ago). Qesem Cave, Israel. J. Archaeol. Sci. 44, 12e21.
Shea, J.J., 2003. The Middle Paleolithic of the East Mediterranean Levant. J. World
Prehist. 17/4, 313e394.
Shifroni, A., Ronen, A., 2000. Observations on the Yabrudian of Tabun Cave, Israel.
In: Ronen, A., Weinstein-Evron, M. (Eds.), Toward Modern Humans: Yabrudian
and Micoquian, 400e50 k-years ago. British Archaeological Reports International Series 850. Archaeopress, Oxford, pp. 119e132.

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005

R. Shimelmitz et al. / Journal of Human Evolution xxx (2014) 1e8

Shimelmitz, R., Kuhn, S.L., 2013. Early Mousterian Levallois technology in Unit IX of
Tabun Cave. PaleoAnthropology 2013, 1e27.
Shimelmitz, R., Kuhn, L.S., Ronen, A., Weinstein-Evron, M., 2014. Predetermined
ake production at the Lower/Middle Paleolithic boundary: Yabrudian
scraper-blank technology. PLoS One 9 (9), e106293. http://dx.doi.org/10.1371/
journal.pone.0106293.

e de
Skinner, J.H., 1970. El Masloukh: a Yabrudian site in Lebanon. Bulletin du Muse
Beyrouth XXIII, 143e172.
Solecki, R.S., 1970. A sketch of the Columbia University archaeological investigations
at Yabroud (Syria). In: Gripp, K., Schtrumpf, R., Schabedissen, H. (Eds.), Frhe
hlau Verlag, Ko
ln, pp. 199e211.
Menschheit und Umwelt. Bo
Stiner, M.C., Barkai, R., Gopher, A., 2011. Hearth-side socioeconomics, hunting and
paleoecology during the late Lower Paleolithic at Qesem Cave. Israel. J. Hum.
Evol. 60, 213e233.
Tsatskin, A., 2000. Acheulo-Yabrudian sediments of Tabun: a view from the microscope. In: Ronen, A., Weinstein-Evron, M. (Eds.), Toward Modern Humans:
Yabrudian and Micoquian, 400e50 k-years ago. British Archaeological Reports
International Series 850. Archaeopress, Oxford, pp. 133e142.
Twomey, T., 2013. The cognitive implications of controlled re use by early humans.
Cambridge Archaeol. J. 23, 113e128.

Valladas, H., Mercier, N., Hershkovitz, I., Zaidner, Y., Tsatskin, A., Yeshurun, A.,
Vialettes, L., Joron, J.-L., Reyss, J.-L., Weinstein-Evron, M., 2013. Dating the Lower
Middle Paleolithic transition in the Levant: a view from Misliya Cave, Mount
Carmel, Israel. J. Hum. Evol. 65 (5), 585e593.
Weinstein-Evron, M., Tsatskin, A., 1994. The Jamal Cave is not empty: recent dis
orient 20, 119e128.
coveries in the Mount Carmel caves, Israel. Pale
Weinstein-Evron, M., Tsatskin, A., Porat, N., Kronfeld, J., 1999. A 230Th/234U date for
the Acheulo-Yabrudian layer in the Jamal Cave, Mount Carmel, Israel. S. Afr. J.
Sci. 95, 186e188.
Weinstein-Evron, M., Tsatskin, A., Weiner, S., Shahack-Gross, R., Frumkin, A.,
Yeshurun, R., Zaidner, Y., 2012. A window into early Middle Paleolithic human
occupational layers: Misliya Cave, Mount Carmel, Israel. PaleoAnthropology
2012, 202e228.
Wobber, V., Hare, B., Wrangham, R., 2008. Great apes prefer cooked food. J. Hum.
Evol. 55, 340e348.
Wrangham, R., 2009. Catching Fire: How Cooking Made Us Human. Basic Books,
New York.
Wrangham, R., Carmody, R., 2010. Human adaptation to the control of re. Evol.
Anthropol. 19, 187e199.

Please cite this article in press as: Shimelmitz, R., et al., Fire at will: The emergence of habitual re use 350,000 years ago, Journal of Human
Evolution (2014), http://dx.doi.org/10.1016/j.jhevol.2014.07.005