Summoned by passersby to a murky canal in western New York State, homicide detectives discover a woman’s partial remains in a suitcase weighed down by a 20-pound bag of rice. Further investigation yields a block of concrete containing her head and a dumpster with still more body parts.

Steven Symes, Ph.D. analyzes saw marks on a
femur using a Leica MZ16A stereomicroscope
that allows for a magnified digital display with
maximum stereo resolution.

What type of weapon had been used to decapitate the victim, cut off her hands, and sever her remaining limbs? The answer could provide valuable clues as police endeavor to put a weapon to a killer, and, ultimately, as prosecutors assemble the evidence needed for a conviction.

Police place a call to forensic anthropologist Steven A. Symes, Ph.D., an authority on knife and saw mark analysis currently lending his expertise to a pioneering new master’s program in forensic and biological anthropology at Mercyhurst College in Erie, Pennsylvania. Symes can discern from the "signature" left on bone what class of sharp-bladed implement might have been used to accomplish a mutilation or dismemberment, be it a serrated knife, a garden-variety treesaw, or a lumberjack-grade chainsaw.

Powerful Evidence
Tool mark analysis has become a highly technical subfield of forensic science. Whether comparing the unique features of bullets, shoe prints, tire tracks, or the class characteristics of saw marks, the accuracy of diagnostic techniquesis vital to strengthening its validity as a legal instrument.

Yet, for all the advances in tool mark scrutiny, particularly in the field of firearms, Symes said few methodologies and no published standards exist for knife and saw mark analysis.

“At one end of the spectrum you have a significant number of cases involving dismemberment and mutilation and the potential for powerful evidence from reliable knife and saw mark analysis; at the other end you have a limited number of forensic practitioners involved in this type of analysis as well as a lack of adherence to Daubert criteria to support their conclusions,” Symes said. “The contradiction is glaring, and poses a growing need in the forensic community.”

Putting Teeth in Saw Mark Analysis
Now, with the support of the Applied Forensic Sciences Department at Mercyhurst College, distinguished colleagues, and funding from The National Institute of Justice,1Symes has undertaken research aimed at establishing a gold standard methodology for knife and saw mark analysis. The goal is to empower qualified forensic scientists to perform reliable analyses and support their conclusions with quantitative and statistical data. The National Forensic Academy andthe American Academy of Forensic Sciences are partners in the project.

Symes’ expertise dates back to 1987 when he began researching what would become a landmark doctoral dissertation on saw mark analysis of bone. Since then, he has provided analysis of cut marks in nearly 200 dismemberment cases and roughly 500 knife wound cases emanating from a wide range of local, state, and national investigative agencies. As a board member of the American Board of Forensic Anthropology, an exclusive group of fewer than 75 board-certified forensic anthropologists in North America, he has also applied bone trauma analysis to human rights investigations in Guatemala, El Salvador, Chile, Indonesia, Mexico, Peru, and Kosovo.

Because saw mark analysis has received little research attention in recent years, Symes maintains that many in law enforcement and forensic science remain unaware of the significant contribution the recognition, interpretation, and documentation of this type of sharp force trauma can add to the resolution of a case. It can narrow the search for a weapon; provide the link, or lack thereof, between a recovered weapon and the crime; and be of enormous evidentiary merit in a court of law.

Unfortunately, with many of these odious cases, valuable clues are often overlooked because many forensic practitioners do not care to include bone as a material within their realm of expertise. Further, organizations not able to sacrifice the time and money necessary for expert analysis are frequently forced to continue their investigations based on little evidence, or worse, inaccurate analysis. Even under expert analysis, the absence of published standards validated by the scientific community can make such evidence vulnerable to derision by defense lawyers as an interpretive art.

Misconceptions Impose Limitations
Historically, at least two major misconceptions contribute to the lack of accuratesaw mark analysis in the forensic sciences. When Symes began his original research two decades ago, conventional wisdom held that every pass of a saw on a bone erased the marks left behind by the previous stroke. But just as people leave fingerprints and firearms produce characteristic defects, he was convinced that blades leave telltale marks of their own.

So, he began by buying and borrowing every imaginable saw – rip saw, hack saw, circular saw, Japanese pull saw, and many more. Using donated arm and leg bones, he made thousands of experimental cuts, attempting pull strokes, false starts, rotary cuts, and others, while under the microscope he studied striations in the groove, or “kerf,” created by the sawing action. In time, he discovered that each saw class did indeed leave its own calling card.

Another misconception that has greatly diminished saw mark analysis in a forensic setting and, in many cases, led to its abandonment as a source of potential evidence is that diagnostic marks are created only when a blade is worn or damaged, thereby resulting in unique characteristics. While the focus on these “defects” is common, all blades leave diagnostic clues as they cut, regardless of wear.

Class Characteristics
Tool mark examiners typically look for “unique” features that produce a positive match. Conversely, knife and saw mark analysis does not necessarily identify a specific weapon, but a “class” of tool, which can be exceedingly useful in the investigative and judicial processes.

As a class of tool, knives can be broken into two categories: single vs. double-edged and serrated vs. non-serrated. When examining a single knife stab wound to a bone, a description beyond those two categories is difficult. If, however, a knife is used to accomplish a dismemberment, the sawing action creates more recognizable patterns. The process can be defined as a “saw dismemberment” and evaluated accordingly, making it possible, for example, to determine features such as “teeth per inch.”

The vast range of saws on the market today can be categorized into approximately 15 classes of tool. The object of saw mark analysis is to recognize characteristics on kerf walls and floors in bone that may accurately reveal size, shape, set, power, and direction of a saw.

  1. Saw size is simply represented by the size of the cut made by individual saw teeth and the combined action of the teeth. Size also reflects the number of teeth per inch (TPI) on a blade.
  2. Saw shape generally applies to the design of the blade, which could be circular, flexible, straight, or a combination.
  3. Set refers to the transverse bending of individual teeth or sets of teeth. Differences in alternating, raker, or wavy teeth influence blade cutting and are recognizable features.
  4. Saw power refers to mechanical vs. hand-powered cutting, both of which produce diagnostic marks.
  5. Direction of cut distinguishes power stroke from passive stroke.


Simplifying the Process
An upshot of this research project is to facilitate a system of saw mark analysis that simplifies an otherwise tedious process, a process Symes may well have evaded himself had a pivotal incident not triggered his realization of its application.

He had just taken a position as a forensic anthropologist in the medical examiner’s office in Nashville, Tennessee in the late 1980s when he was handed three unrelated dismemberment cases. A detective on one of them pointed to a gash in a bone and asked Symes what he made of it. He answered that it looked like “a saw mark in an ulna.” But, it was not the answer the detective was looking for and he told Symes as much: “I know it is a saw mark. But, you’re the bone doc. What kind of saw mark?”

Suddenly, Symes realized that his meager insight was of little value to the detective’s investigation, but if he could pin-point at least a class identification of the weapon thathad been used in the dismemberment, he might provide a lead to follow.

Eventually, his research allowed him to discern far more than “a saw mark in an ulna.” He could, for example, show that a particular dismemberment could only have been accomplished by a hand-powered, reciprocating motion saw, consistently narrow like a bow saw or Japanese pull saw, with 18 teeth perinch set in an alternating pattern.

Symes’ latest research is aimed at enabling other forensic practitioners to reach similar conclusions with a degree of statistical certainty, to do it in-house, expeditiously, and at minimal expense. His team includes Anne M. Kroman, MA, a doctoral candidate in physical anthropology at the University Tennessee, Knoxville; and Susan M. Thurston Myster, Ph.D., associate professor of anthropology at Hamline University in . Paul, Minnesota. Other contributing personnel include Dennis Dirkmaat, Ph.D., DABFA, director of applied forensic sciences at Mercyhurst College; Luis Cabo-Perez, Mercy-hurst forensic science lab director; Stephen Ousley, Ph.D., director of the Repatriation Osteology Laboratory, National Museum of Natural History, Smithsonian Institution; andtwo Mercyhurst graduate students, Christopher W. Rainwater and John J. Matia.

Plan of Action
The project will:

  • Utilize the vast range of saws and knives on the market today to make experimental cuts in human bone and describe key variables useful in analysis. This phase builds on Symes’ original research.
  • Complete a validation study of the variables to ascertain which offer the most utility in correctly identifying knife and saw marks and assigning them to the appropriate class of tool. As part of this phase, input will be sought from crime labs, forensic tool mark examiners, anthropologists, and general practitioners across the United States and abroad.
  • Develop standards for assessing saw mark characteristics. In order to provide a methodology that meets criteria for standards for each variable and the criteria for evaluating each variable need perfecting. For example, there are several techniques that can be employed to calculate the number of teeth per inch (TPI) of a saw. Symes’ team plans to evaluate each of these methods for accuracy, ease of use, repeatability, and reliability.
  • Establish a documentation standard for cut marks in bone in the forensic setting. Better documentation before burial or cremation will foster accurate recovery of tool mark evidence.
  • Produce a comprehensive user-friendly handbook that describes diagnostic knife and saw mark characteristics, analytical techniques, documentation procedures, and reporting protocol.
  • Create a computer program as a complement to the handbook that provides a statistical approach to estimate the probability of tool matches based on variable input. The program design would be similar to FORDISC 3.0, an anthropological program that provides statistical support for metric classification of human skeletons. The enhanced statistical scrutiny is intended to augment the use of saw mark analysis results in a court of law by allowing forensic practitioners to report probabilities and error rates.

“There are broad implications for this research in terms of knowledge, policy, and the everyday practice of law enforcement,” Symes said. “The NIJ grant facilitating this study is the manifestation of a new era of government funding designed to advance the analysis and interpretation of evidence from criminal acts. In these days of highly publicized forensic science, we seethis funding as vital to validating the judiciary benefit of tool mark evidence.”

1. This project is supported by Award No. 2005-IJ-CX-K016, awarded by the National institute of justice, Office of Justice Programs, US Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication/program/exhibition are those of the authors and do not necessarily reflect the views of the Department of Justice.

Deborah W. Morton is a former newspaper reporter, freelance writer, and currently serves as assistant director of publications at Mercyhurst College. She canbe reached at

Photos by Christopher W. Rainwater and Steven A. Symes, Ph.D.

From the case files of Steven A. Symes,Ph.D.

On the surface, it’s just another box of bones, but it’s what lies within those bones that can breathe new life into an investigation.

As I inventory the contents of a package recently sent to me by a Minneapolis medical examiner working a homicide and dismemberment case, I find sections of neck, arm, and leg bones that had been harvested as evidence from the autopsy. Dismemberment frequently involves severing the head and limbs from the torso, whether to disguise or to ease transport, so this assortment of bone specimens is typical of what I would be asked to analyze in a dismemberment case.

My challenge as a forensic anthropologist specializing in trauma to bone is to discover what class of sharp-bladed implement was used to accomplish the dismemberment. While my findings may not speak to the murder per se, such ancillary information can increase the success of both the criminal investigation and the trial in determining the facts of the case.

Inventory: Bone specimens from dismemberment case inventoried for tool mark analysis.

As I begin my examination of the tool marks, I find class characteristics immediately. The analysis of saw marks begins with a “kerf,” which is the cut in bone produced by toothed blades. A kerf consists of walls and a floor. Already, this case reveals an interesting floor. Instead of finding the typical squared-off U-shaped trough, I discover cut marks resembling a “W” (Figure 1). Immediately I am able to narrow the description of the tool used in this crime by the design of its blade. It is a saw that is hand-powered and has filed teeth.

Figure 1: Views of left femur show “W” shape created by saw action.

Filing saw teeth is a technique designed to make carpenter saws efficiently cut across the grain of wood. This alteration, which produces the “W” cut mark, brings a tooth to a point, so cutting is actually accomplished with tiny knife-shaped teeth instead of chisels. Filed teeth earn the tool a classification of “crosscut saw.”

Saws without filed teeth are traditionally called “rip saws,” and with today’s mass production of saws, most feature the “rip” tooth design. Since mechanically powered saws typically cut with quick, powerful action, the crosscut design is not usually desired, whereas the sturdier rip-designed teeth are common to power saws.

Now, as I continue to narrow the search for a dismembering tool, I measure the minimum kerf width and determine it to be 0.05 inches (Figure 2). That determination rules out the hefty power saws and the narrow bow saws. The tool used in this crime is somewhere in between, more in keeping with a carpenter saw.

Figure 2: Views of right femur illustrate minimum kerf width of 0.05 inches.

Next, in observing the shallow kerf of an apparent false start, I see that four teeth had penetrated the bone when the saw was interrupted in mid-stroke and removed. The false start reveals troughs within the larger kerf (Figure 3), indicating several key diagnostic features, including teeth per inch (TPI), a universal way of describing saw size; and direction of saw. From the striations created by the false start, I conclude that the saw has nine teeth per inch set in an alternating pattern and cuts on the push stroke.

Figure 3: False start in right femur reveals key diagnostic features, including teeth per inch.

To summarize, my analysis at this point has uncovered the following basic characteristics of the suspect tool:

  • hand-powered
  • crosscut design
  • cuts on the push stroke
  • nine TPI
    teeth set in an alternating pattern
  • minimum kerf width of 0.05 inches.

Those are certainly enough features to guide an investigator in the search for the suspect tool, or to inform a jury in the event a tool is recovered. My job would seem to be all but finished.

But, wait, the left femur throws me a curve. There among the kerfs I have grown accustomed to seeing, a different shape emerges. There are “V”s among the “W”s. The V kerf shape indicates that the blade tapers to a single beveled edge, which is consistent with a knife. Further, these cuts are all shallow, suggesting a weapon of less efficiency.

Knives used as saws work on the same principle as classic carpenter saws. They simply lack the alternating set to the teeth that enables a blade to cut hard material without binding.

Because I expect some variation in the marks produced by an individual tool, I am reluctant to endorse the presence of another tool without conclusive findings. In this case, there is undeniable evidence that two classes of tool have been used to saw into the bones (Figure 4). My analysis of the “V” kerf shape reveals the second tool as a fine-toothed, serrated knife with teeth situated on the blade’s left side.

Figure 4: Left femur reveals multiple shallow cuts from knife action and larger saw cuts.

Why the second tool? It is common for a perpetrator to use a knife to cut the soft tissue and expose the bone for the saw. What is interesting in this case, however, is that the knife marks are consistent with frustrated attempts to cut into the bone, rather than create superficial cuts to clear the way for a more efficient tool. In the end, the perpetrator appears to have abandoned the knife for the saw to complete the dismemberment.

As investigators continue their search for a murderer, my analysis of the dismemberment aspect of this case increases their options by providing additional forensic evidence that is accurate, precise, and reliable: They are looking for two classes of tool: first, a hand-powered, crosscut saw that cuts on the push, with nine teeth per inch set in an alternating pattern; and secondly, a fine-toothed, serrated knife with teeth on the left side of the blade.

Not bad for a day’s work with a box of bones.

Steven A. Symes, Ph.D., DABFA, is faculty at the Mercyhurst Archaeological Institute at Mercyhurst College in Erie, Pa., where he lends his expertise to a new landmark master’s program in Forensic and Biological Anthropology and an undergraduate program in Applied Forensic Sciences. In addition to teaching, he has published extensively, lectured worldwide, and served as a consultant to pathologists, attorneys, and criminal investigators in numerous states and Canada, imparting his expertise in skeletal trauma, dismemberment, and mutilation.