LITHAN

1
LITHAN

• LITHic Analysis of stone tools

2
What is LITHAN?

• System to identify the type of stone tool
• Created 20 years ago by Roger Grace
• Programmed using Macintoshs Hyperstudio
• At the time it was tested by a group of students
• The students all came out with similar answers,
  even when some of the values entered were
  different

3
What does it do?

• LITHAN is a set way of identifying tools
• Also provides information on
• Blank type ( blade, bladelet, flake, chip,
  fragment or chunk)
• Flint knapping Technology ( blade, flake,
  Levallois)
• Hammer mode ( soft or hard )
• Cortex

4
How does LITHAN work?

• Takes 19 different attribute values
• Compares these values to the system rules
• Produces a screen with the conclusions on it

5
Picture of the main page
6
Attributes

• Two types of attributes used
• Measured
• Observed
• Measured values measured, normally using
  callipers
• Observed observed by the naked eye or using a
  low power microscope

7
Measured Values

• 7 measured attributes
• Tool length
• Tool width
• Tool thickness
• Mid-point width
• Mid-point thickness
• Platform width
• Platform thickness

8
Observed Attributes

• 12 different observed attributes
• These normally have a screen with specific
  options to select from
• These allows only certain values to be entered

9
Example Picture of the cards
10
Observed Attributes

• Platform type
• Lateral edges
• Dorsal ridges
• Cortex
• Percussion
• Butt
• Bulb
• Retouch
• Position of retouch
• Retouch type
• Edge form
• End form

11
Special Options

• If certain options are entered a screen
  automatically pops up to clarify observations and
  help with identification
• Platform type core
• Retouch type burin blow
• Tool type arrowhead
• Tool type microlith

12
Core

• Cards to clarify options when Platform type Core

13
Burin

• Cards to clarify when Retouch Burin Blow

14
Arrowhead

• Card used when Type arrowhead

15
Microliths

• Card used when type microlith

16
Rules

• LITHAN has 30 pages of programming covering the
  rules in the system
• Has two types of rules
• General applied all the time
• Secondary applied only if a specific conclusion
  is previously reached

17
Examples

• Rule to determine blank type
• If (length/width) gt 2 and width lt 12 mm
• Then Blank type Bladelet
• Rule to determine technology
• If Platform thickness lt 5 and Platform type
  prepared and Lateral Edges parallel and
  Dorsal ridges parallel
• Then Technology type Blade
• Rule to determine hammer mode
• If percussion no cone and Butt un-lipped
  and Bulb diffuse
• Then hammer mode soft

18
Examples

• Rule determining type of tool
• If (length width) gt 0 and retouch position
  distal
• Then type End scraper
• Secondary rule
• If type End scraper and End form Round
• Then type End scraper
• Else if type End scraper and End form
  Carinated
• Then type Carinated End Scraper

19
FAST

• Functional Analysis of Stone Tools

20
FAST

• Functional Analysis of Stone Tools
• Sister program to LITHAN
• LITHAN results can be used in the program
• FAST uses the information it gathers to determine
  the wear on an artefact
• Wear meaning what sort of material it was used on
• e.g.. Fish scale, hide, antler, bone, etc

21
How does FAST work?

• Takes in 20 different attribute values
• Uses fuzzy logic to calculate the best possible
  answer
• Each attribute has allowable values which are
  selected from a new card by the user

22
Main Page
23
Attributes

• Three types of attributes
• Measured
• Macro Observed
• Observation using low powered magnification
• Micro Observed
• Observations made under high magnification

24
Measured

• Edge angle
• Length
• Thickness
• Profile
• Shape

25
Macro Observed

• Gloss on the Edge wear
• Fractures a count of them
• Fracture type
• Rounding is the edge of tool rounded

26
Micro Observed

• Fractures and Fracture Type
• Rounding
• Micro topography
• Distribution
• Distribution type
• Invasiveness
• Linear features
• Striations
• Polish development

27
Attributes

• Some absent values can be used to determine use
• E.g.. If the edge is not round, tool was probably
  used on a softer material
• Other values, while present, are not important
  for the determining of use
• E.g.. Micro fractures are not diagnostic because
  they can occur from any sort of contact motion

28
Results

• Once attributes are entered they are displayed on
  two cards
• One for the Measured and Macro Observation
• One for the Micro Observations

29
Macro and Measured
30
Micro
31
Rules

• 3 types
• General
• Fuzzy
• Function

32
General Rules

• Attributes are used singlely or in combination
• E.g.. (edge anglelt 30) then Angle cutting
  soft material
• E.g.. (fractures absent) and (edge angle gt 30
  and lt60) then Angle medium material
• Conclusions reached by the rules are Observations
• This Observations are then used in the Fuzzy Rules

33
Fuzzy Rules

• The fuzzy rules use the conclusions from the
  general rules
• Each conclusion have specific numerical value
  placed on them according to the rules
• The values are totaled to produce a score value
• The score value is used in the function rules

34
Example of Fuzzy Rules

• Each variable (egg. Angle, length, profile) count
  as 2 (see the Macro Card)
• If the surfaces (Ventral/Dorsal) have values the
  add 0.5
• Unless retouch, no polish, or no effect
  Then add 1 because they are more telling
• Non-Diagnostic 0
• All motion scores (except rotational) are doubled
  

35
Functional Rules

• Use the Score values to determine
• Function
• Hardness of material
• Type of Material

36
Example of Functional Rules

• Function
• E.g.. cutting lt 4 and scraping gt 8 and
  grooving lt2 and whittling lt 2 then function
  scraping
• Hardness
• E.g.. (soft gt 4 and lt 8) and (medium gt0 and lt
  2) and hard 0 then hardness soft
• Type
• E.g.. soft lt 6 and medium gt 5 and hard 0
  then type wood
• Type can also have more complex rules involving
  motion, and other values

37
Displaying Results

• Results are displayed on a single screen
• If unable to determine a value then Insufficient
  Data is displayed
• Occurs in 2 cases normally
• Not enough use wear on tool
• Use wear not consist with a specific use

38
Result Card
39
Analysis of FAST and LITHAN

• LITHAN determines type of tool
• The value found by LITHAN and be put into the
  FAST system
• FAST finds the use of the tool
• Both systems tested at Universities
• FAST has a 90 success rate during tests
• LITHAN tests have users coming up with the same
  answers even when some value are different for
  the same tool

40
General Advantages

• Rules are cemented and written down
• Different people will still get the same values

41
General Disadvantages

• If rule is determined to be incorrect (e.g. Value
  given more weight then it should have) all
  findings have to be suspect

42
Advantages of LITHAN

• Different people will get the same result
• Allows comparison without worry of results being
  different because of the typologist
• LITHAN does not use Fuzzy Logic so the answers it
  provides are presumed accurate
• If not enough information is provided (i.e..
  Because tool is broken or pieces are missing)
  then the value is designated indeterminate
• Can update as new rules are found
• Records the tool type into a permanent record

43
Disadvantages of LITHAN

• Does not provide a great deal of useful
  information
• So its a scraper/arrowhead/other tool, big deal,
  most past societies have these
• Most of the conclusions can be drawn from eye
  sight

44
Advantages of FAST

• Runs a complex analysis quickly once values
  entered
• Information is relevant to research
• E.g.. Scraping hide implies hunting

45
Disadvantages of FAST

• Uses fuzzy logic, so have to be aware the answer
  is weighted

46
Observations

• When designing the programs the creator had to
  think about three things
• Facts to include
• Rules to include
• Storage of solutions

47
Observations

• The facts
• all contribute to the solution (e.g. colour of
  tool tells nothing of type or use of tool)
• If too few no accurate answer gained
• If too many spend so much time entering data,
  easier to do it by hand

48
Observations

• The rules
• if wrong or too vague then the system does not
  provide an accurate answer
• Have to provide for all possible cases
• Must be updated if new information gained (e.g.
  FAST has a lot of information about use wear as
  it is applied to fish scales because it was used
  in a Norwegian study that had fish scale wear on
  tools)
• Storage
• Have to store answers for reference
• Dont want storage to be hard to search
• Both LITHAN and FAST use Hypercard for this

49
Conclusion

• LITHAN is a system that is useful in a limited
  area
• Helps with the consistency of identification of
  tools
• FAST is a system with a wider range of usefulness
• LITHAN while interesting is not helpful, other
  than in a data gathering way like a database
• FAST provides important information and shows the
  reasoning behind a complicated process
• If the program the systems are written on was
  updated the programs would be more useful

50
Conclusion

• LITHAN and FAST use some of the same attributes
• LITHAN has more in depth measurements of tool
• FAST has micro observations
• Might combine the 2 programs to run as one
• LITHAN has a limited value, FAST is more useful
  for studies
• Would require the reworking of the rules

51
Bibliography

• LITHAN Movie by Roger Grace
• http//web.mac.com/rgrace2/ES/Movie.html
• LITHAN Grace, Roger
• http//www.hf.uio.no/iakh/forskning/sarc/iakh/lith
  ic/expsys.htmlanchor130219
• FAST Grace, Roger
• http//www.hf.uio.no/iakh/forskning/sarc/iakh/lith
  ic/FAST.html