The use of fire was a key factor
in the evolution of Homo sapiens, not only for the creation of more
sophisticated tools but also for making food safer, which in turn aided brain
development.
To date, only five sites with
fire evidence dating back 500,000 years have been found worldwide, including
Wonderwerk Caves and Swartkrans in South Africa, Chesowanja in Kenya, Gesher
Benot Ya'aqov in Israel, and Cueva Negra in Spain.
Now, an Israeli research
team has used artificial intelligence algorithms to discover a sixth site that
shows traces of human fire! The study revealed evidence of human use
of fire at a late Paleolithic site in Israel. The research results have
been published in the journal PNAS.
AI forays into archaeology
Traditional archaeological
methods for the identification of fire sources at early hominin sites rely
primarily on the visual assessment of altered sediments, cuttings, and bones,
such as soil reddening, discoloration, warping, cracking, shrinkage, darkening,
etc., etc., which may underestimate how common fire was in humans at the time.
In this study, the authors'
team developed a spectroscopic "thermometer" based on Raman
spectroscopy and deep learning algorithms to estimate thermal exposure to
flint artifacts and detect the atomic structure of distorted materials at
extremely high temperatures, thereby compensating for the use of Possible absence
of fire traces in visual features.
Studies have shown that the early
Paleolithic open-air site (Evron Quarry) in Israel contains remains of burned
animals and debris, dating between 1 million and 800,000 years ago.
 |
| Legend: From left to right are Filipe Natalio, Ido Azuri, Zane Stepka |
The research team first examined
material excavated at Evron Quarry in 1976-1977 and found no visually obvious
evidence of heat-related features such as reddening of soil, discoloration or
cracking of flint tools, shrinkage, or discoloration of animal remains.
 |
| Caption: The archaeological excavation site of the Evron Quarry site |
The team tested many
approaches, including traditional data analysis methods, machine learning modeling,
and more advanced deep learning models. Popular deep learning models
have specific architectures that outperform others, and the benefit of using AI
technology is that it can analyze the chemical composition of materials and use
this to estimate their thermal exposure.
AI technology can reliably
distinguish whether modern flint has been burned, and it can also reveal the
temperature at which it burned. The heat of the fire can cause changes in
nearby stones, and burning changes bone structure at the atomic level, with
corresponding changes in the infrared spectrum.
In this study, the team used a
deep learning model (a one-dimensional convolutional neural network) to learn
the Raman spectral patterns of flint artifacts to estimate the temperature of
stone tools. Compared to a fully connected artificial neural network
(FC-ANN), the model performed better, reducing the mean absolute error between
the true and estimated temperatures from 118 °C to 103 °C.
First, the team pre-trained
modern flint collected from various locations in Israel and heated to a
known temperature under laboratory-controlled conditions. Second, the
trained model was applied to unknown samples (ie, stone tools collected from
the Evron Quarry site). The team used a supervised deep learning approach
to correlate Raman spectroscopy with the heating temperature of the
flint. This approach relies on irreversible thermally-induced structural
changes that occur in the organic and inorganic components of flint while
overcoming its inherent variability. The advantage of using a deep
learning model for temperature estimation is that it can approximate any
nonlinear decision boundary between heat and spectral changes due to heat in
alpha-quartz, moganite, and the D and G-band spectral regions.
In the image below, the stones
do not visually show any signs of being burnt, but by using a deep
learning model to estimate the thermal exposure of the UV Raman spectra
collected from the stones, it was found that they were all heated between 200°C
and 600°C. This suggests that ancient humans could control
fire rather than just use natural wildfires.
follow-up discussion
For the excavated bones,
the research team also experimentally confirmed that they had been burned
by fire. Chazan, one of the authors, said: "Without AI-validated
flint results, no one would bother to test the heat exposure of these
bones."
The study, though, has not been
able to determine whether the site's tools were burned by natural or artificial
fire. The spatial variation caused by burning traces can be interpreted as
evidence of human intervention since natural fires often result in homogeneous
thermal changes across the burning area.
The authors acknowledge that
wildfires and uneven vegetation can also contribute to uneven temperature
distribution across the region and that temperature is not a reliable
criterion to distinguish between wildfires and artificial fires. Still,
the estimated temperature of Stone Age utensils and the presence of burned
fauna suggest the possibility that ancient humans at the site used fire.
In the future, the methods used
in this study can be extended to other late Paleolithic sites, which will
potentially expand the spatiotemporal understanding of the relationship between
early hominins and fire, opening windows into early human life.
Reference link:
https://www.pnas.org/doi/full/10.1073/pnas.2123439119
https://news.sciencenet.cn/htmlnews/2022/6/480888.shtm
https://www.timesofisrael.com/old-flame-israeli-researchers-find-evidence-of-fire-use-nearly-1-million-years-ago/
