Chapter 10: Blood and Body Fluids: Serology, Presumptive Testing, and Bloodstain Pattern Analysis

"Blood is a chatty witness, but it does not always say what we want to hear — and an interpreter who hears only what the detective wants has stopped listening." — constructed teaching epigraph, in the voice of this book

Overview

Stand in a room where someone has bled and you are looking at two completely different questions wearing the same red color. The first is a chemistry-and-biology question: is this blood, is it human, and whose is it? The second is a physics question: what happened here — and can the shapes these drops left behind tell me? The first question, asked carefully, sits near the strong end of forensic science, because it hands off to DNA (Chapters 7–9), the one method the field has truly validated. The second question is where this chapter earns its keep as a cautionary tale, because reconstructing events from the geometry of stains — bloodstain pattern analysis — is far more contested, far more examiner-dependent, and far more dangerous in court than any television drama has ever admitted.

This is a chapter about telling those two questions apart and never letting one borrow the credibility of the other. We begin with serology — the identification and grouping of biological fluids — and the workhorse distinction between a presumptive test, which screens fast and cheap but can be fooled, and a confirmatory test, which costs more and proves more. We will run the chemistry of the classic blood screens honestly: how Kastle-Meyer turns pink, why luminol glows blue in the dark, and exactly what each one false-positives on, because the false positives are the whole story of where these methods sit on the validity spectrum. Then we follow the stain to its real destination: the moment serology hands the swab to the DNA analyst and steps back, because "it is human blood" and "it is his blood" are separated by an entire laboratory.

Only then do we turn to the geometry — droplet size and shape, directionality, angle of impact, the area of origin reconstruction that strings a room with lines to a point in space — and we hold it to the same merciless yardstick we hold everything else. The 2009 National Academy of Sciences report singled bloodstain pattern analysis out for unusually blunt criticism, and the courts have since produced both convictions and exonerations turning on disputed pattern testimony. By the end you will be able to read a stain for the modest, defensible facts it carries, and to recognize the exact moment an analyst quietly inflates "consistent with" into "this is how the murder happened."

In this chapter, you will learn to:

• Define serology and tell a presumptive test apart from a confirmatory test, naming what each can and cannot establish.
• Explain how Kastle-Meyer, leucomalachite green, and luminol detect blood — and recite the specific false positives that fix each one's place on the validity spectrum.
• Explain why "is it human, and whose is it?" — not "is it blood?" — is the question that decides relevance, and how serology hands off to DNA.
• Read a bloodstain pattern by size, shape, directionality, and angle, and state the limited inferences a pattern can honestly support.
• Describe the area of origin reconstruction and the error sources that make it an estimate, not a measurement.
• Locate bloodstain pattern analysis (BPA) on the NAS/PCAST validity spectrum and state what a BPA analyst may honestly say on the stand.

Learning Paths

🔎 Investigator/CSI: Your work decides whether any of this is possible. Sections 10.2 and 10.3 — presumptive screening at the scene, photographing and collecting a stain before it is consumed, and packaging it so the DNA lab can still use it — are the heart of your job. Document the pattern (10.4) before you touch it; you cannot un-disturb a scene. 🧪 Lab analyst: Weight 10.1–10.3. The presumptive/confirmatory logic, the false-positive chemistry, and the serology-to-DNA handoff are your bench. Know precisely what your screening test does and does not license you to write in a report. ⚖️ Law/courtroom: Sections 10.5 and 10.6 are where the cross-examination lives. BPA is among the most attackable forensic testimony in the building; learn the verbs an analyst may and may not use, and the NAS finding that frames every challenge. 👥 General reader/juror: A confident analyst recreating a murder from blood spatter is gripping television and shaky science. Sections 10.4–10.6 are the antidote: what the stains can really tell you, and what is the analyst's story rather than the blood's.

10.1 Identifying body fluids: presumptive vs. confirmatory tests

Before any of the glamorous reconstruction, a stain on a carpet is just a stain. The first forensic question is humble and decisive: what is it? The discipline that answers it is serology — the forensic identification, characterization, and (historically) grouping of blood and other body fluids such as semen, saliva, urine, sweat, and vaginal fluid. The name comes from serum, the fluid fraction of blood, but the field long ago expanded to every fluid the body can shed at a scene. In the era before DNA, serology was also the individualizing science — analysts typed ABO blood groups and protein markers to narrow a stain to a fraction of the population. DNA has since taken that job almost entirely (we will see the handoff in 10.3), and modern serology has narrowed back to two questions it answers well: is a biological fluid present, and which one is it?

The logic that organizes all of serology — and much of the chemistry chapters to come (Chapters 20–24) — is a two-stage funnel. You screen fast, then you confirm slow.

A presumptive test is a fast, sensitive, inexpensive screening test that indicates a substance is probably present and, just as importantly, indicates where it is not. Presumptive tests are designed to be over-eager: they would rather flag a stain that turns out to be rust than miss a stain that turns out to be blood. Their value is triage. A CSI cannot DNA-test an entire kitchen; a presumptive test tells them which three of two hundred reddish marks are worth collecting. The price of that sensitivity is false positives — substances other than the target can trigger the reaction — which is why a presumptive result is never, by itself, an identification.

A confirmatory test is a more specific test that establishes, to a high degree of certainty, that the substance is the target — that this is human blood, or that this is semen, and not merely something that behaves like it in a screening reaction. Confirmatory tests are typically slower, costlier, sometimes consume more sample, and answer a narrower question, but they carry the weight a presumptive test cannot.

        THE TWO-STAGE FUNNEL OF SEROLOGY
        (the same logic runs through all of forensic chemistry)

   many candidate stains
   ─────────────────────►  [ PRESUMPTIVE TEST ]  ─── negative ──►  excluded (reliably)
        (fast, cheap,            │
         sensitive)              │ positive ("could be")
                                 ▼
                        [ CONFIRMATORY TEST ]  ─── negative ──►  not the target after all
        (specific,               │
         identifies)             │ positive ("it is")
                                 ▼
                        [ DNA ANALYSIS ]  ──►  whose? (Chapters 7–9)
                                              ── source attribution lives here, not in serology ──

Read that diagram in both directions, because the negative arrows are quietly the most powerful. A clean presumptive negative is one of those reliable exclusions the book keeps returning to (Theme 1): a properly performed Kastle-Meyer test that stays colorless tells you, with real confidence, that detectable blood is not there. Forensic science's surest voice, again, is the one that says no. A presumptive positive, by contrast, says only "maybe — keep going." The funnel never lets a screening result masquerade as an answer.

🔬 At the Bench The order of operations matters as much as the chemistry. You presumptively test a small portion — a swab rubbed on the edge of a stain, a cutting from a fabric — never the whole stain, because the next step may need pristine material for DNA. You photograph and document the stain before you sample it. And you run controls every time: a positive control (a known sample of the target, to prove your reagents work) and a negative/substrate control (the clean material next to the stain, to prove the surface itself isn't causing the reaction). A presumptive test reported without controls is an anecdote, not a result.

For body fluids beyond blood, the same funnel applies with different reagents. Semen is screened presumptively for the enzyme acid phosphatase (an AP test, fast color change) and confirmed by visualizing spermatozoa under a microscope or by detecting the protein p30/PSA (prostate-specific antigen), which matters in cases involving vasectomized or azoospermic men who produce no sperm. Saliva is screened for amylase. Urine, sweat, and feces have their own markers. Newer RNA-based and immunochromatographic ("strip") tests can identify the fluid type more specifically and are gaining ground, though their courtroom acceptance is still maturing. The principle is constant across all of them: screen, then confirm, then — only then — ask whose. Throughout this chapter, hold onto that ordering. Most serology errors are not chemistry failures; they are an analyst, or an attorney, treating a presumptive flag as if it were a confirmed identification.

10.2 Blood: Kastle-Meyer, luminol, and their false positives

Blood is the fluid forensic science cares about most, and the one whose presumptive chemistry you should understand in detail — not because the reactions are hard, but because their false positives are the entire argument about where these methods sit on the validity spectrum.

Every common blood screen exploits the same trick. Hemoglobin, the oxygen-carrying molecule that makes blood red, contains a heme group with an iron atom, and heme behaves like a peroxidase — an enzyme that catalyzes reactions involving hydrogen peroxide. The tests supply a colorless (or non-luminescent) molecule plus hydrogen peroxide; if heme is present, it speeds an oxidation that produces a vivid color or a glow. The reactions do not detect "blood" as such. They detect peroxidase-like activity — and that is precisely the loophole that lets non-blood substances fool them.

Kastle-Meyer is the classic. Named for the chemists who developed it in the early twentieth century, it uses phenolphthalein (in a reduced, colorless form) plus hydrogen peroxide. The CSI rubs a moistened swab on the edge of the stain, adds the phenolphthalein reagent, then adds the peroxide. If blood is present, a bright pink color appears within seconds. The timing convention matters and is part of the method's discipline: color that appears only after the peroxide is added is the meaningful result; color that appears before the peroxide points to a different chemistry and is not counted as a positive. Kastle-Meyer is exquisitely sensitive — it can flag blood diluted thousands of times — and that sensitivity is both its power and its weakness. A close cousin, leucomalachite green (LMG), works the same way but turns blue-green; a third, Hemastix, is a clinical urine-dipstick strip repurposed for field screening.

🔬 At the Bench A Kastle-Meyer "positive" is a presumptive result, full stop. It tells you the swab carries peroxidase activity consistent with blood. It does not tell you the blood is human (a horror-movie amount of it could be from a deer field-dressed in the garage), it does not tell you whose blood it is, and it does not tell you when it was deposited. The honest report line is: "Stains tested presumptively positive for blood; confirmatory species testing and DNA analysis were requested." Anything beyond that, written from a Kastle-Meyer result alone, overstates what the chemistry can carry.

The famous member of the family is luminol — the reagent that makes the darkened-room scenes television loves. Luminol is sprayed as a solution; in the presence of heme (and an oxidizer), it undergoes chemiluminescence, emitting an eerie blue glow visible only in darkness. Luminol's gift is visualization of the invisible: it reveals latent (cleaned-up) bloodstains, diluted bloody footprints, drag marks, and spatter that the naked eye cannot see, even after a scene has been wiped or washed. An offender who mops up a pool of blood may leave a luminol-reactive ghost of the original pattern across a whole floor.

But luminol's powers come with serious costs that an honest analyst states out loud:

• It is presumptive, not confirmatory. Luminol glows for blood, but it also glows for a notorious list of innocent substances.
• The false-positive list is long and ordinary. Luminol reacts with certain metal ions (copper, iron), with bleach and other oxidizing cleaners (which is a cruel irony: cleaning a scene with bleach can produce a strong, blood-mimicking glow), with some plant peroxidases (notably horseradish), and with substances like rust and some paints. A glow is a question, not an answer.
• It can damage DNA. Luminol formulations and the alkaline conditions they use can degrade the very DNA you hoped to recover, so it is deployed thoughtfully, often after other documentation, and the lab accounts for its effect on downstream typing.
• It illuminates, then vanishes. The glow is transient and must be photographed in real time with long exposures in darkness — there is no second look.

   BLOOD PRESUMPTIVE TESTS — WHAT THEY ACTUALLY DETECT, AND WHAT FOOLS THEM
   (all exploit heme's peroxidase activity → none is specific to blood)

   TEST              SIGNAL            FALSE-POSITIVE SOURCES (illustrative, not exhaustive)
   ───────────────   ───────────────   ──────────────────────────────────────────────────
   Kastle-Meyer      pink              plant peroxidases (horseradish, some roots),
   (phenolphthalein)                   certain metal salts, some chemical oxidizers
   Leucomalachite    blue-green        same families as above
   green (LMG)
   Luminol           blue glow (dark)  bleach & oxidizing cleaners, copper/iron ions,
                                       horseradish & other plant peroxidases, rust, some paints
   ─────────────────────────────────────────────────────────────────────────────────────
   POSITIVE  = "peroxidase activity consistent with blood — CONFIRM and DNA-test"
   NEGATIVE  = "no detectable blood here" (a reliable exclusion — the powerful result)

So how do you confirm that a presumptive-positive stain truly is blood, and human blood? Two further steps. Confirmatory tests for blood — historically crystal tests such as the Takayama (hemochromogen) and Teichmann (hemin) tests, which form characteristic microscopic crystals from heme — establish that the substance is blood. Species determination then establishes it is human blood: classically the precipitin (Ouchterlony) test, which uses human-specific antibodies, and today most often the DNA analysis itself, which only amplifies human (or targeted) DNA and so answers "human?" and "whose?" together. Note the elegant collapse here: modern DNA testing has folded the old "is it human?" confirmation and the old "whose is it?" individualization into a single workflow, which is one reason classical serology has shrunk.

⚠️ Junk-Science Alert The dangerous overstatement with luminol is not the chemistry — it is the narrative. An investigator who finds luminol reactivity in a wiped-down bathroom and testifies that "the defendant cleaned up a bloodbath" has leapt from a presumptive, non-specific glow to a story about a person's actions. The glow could be old blood from a nosebleed, a cut while shaving, a butchered chicken — or bleach, or copper plumbing residue, or horseradish in the drain. Luminol can direct you to test something; it cannot conclude anything by itself, and it certainly cannot date the stain or name its source. When you hear "luminol lit up the whole room," the next question is always: confirmed as human blood, by what test, and attributed to whom?

Where do these tests sit on the validity spectrum (Theme 2)? Favorably, when used honestly. Blood detection by presumptive screening followed by confirmatory species testing and DNA is well within the strong-to-moderate range: the chemistry is understood, error modes (the false positives above) are known and named, and the confirmatory backstop is rigorous. The validity problem with blood evidence almost never lives in the chemistry. It lives one step later — in the pattern interpretation we reach in 10.4, and in the narrative an analyst hangs on a glow. Keep detection and interpretation in separate mental boxes. Confusing them is the chapter's central error.

🔍 Check Your Understanding 1. A bathroom tile gives a strong luminol glow. Name three innocent explanations other than human blood, and the single follow-up test that would resolve whether it is blood at all. 2. Why is a Kastle-Meyer negative a more confident statement than a Kastle-Meyer positive? (Connect this to Theme 1.)

10.3 From stain to source: where serology hands off to DNA

Here is the most important sentence in the first half of this chapter: serology tells you what a stain is; only DNA can tell you whose it is. The entire value of careful collection in 10.1–10.2 is that it preserves the stain well enough for the analysis in Chapters 7–9 to answer the question that actually decides cases. A stain confidently identified as human blood but unattributable to anyone is, in many cases, almost worthless — and a stain attributed to the wrong person, through contamination at collection, is worse than worthless. The handoff between serology and DNA is therefore not a formality. It is where biological evidence is most often won or lost.

Recall from Chapter 7 that blood is one of the richest sources of nuclear DNA at a scene, because it is full of white blood cells (red cells, famously, carry no nucleus and no nuclear DNA — a point that trips up newcomers). A bloodstain the size of a dime can yield a complete single-source profile. But that potential is fragile. Everything that protects DNA — and every way it can be destroyed — funnels through the seconds when a CSI decides how to collect a stain.

🔬 At the Bench The collection drill for a dried bloodstain is a small, exact ritual, and it is the same ritual that makes the cold case's doorframe stain (below) admissible: 1. Document first. Photograph the stain in place, with and without a scale, before any contact. You cannot recollect a pattern you have smeared. 2. Presumptively test a small edge — never the center, never the whole stain. 3. Collect by the right method. A dried stain on a non-absorbent surface is collected by the wet-then-dry swab technique: a sterile swab moistened with sterile water lifts the stain, followed by a dry swab to capture the rest. A stain on fabric may be cut out (with a control cutting of clean adjacent fabric). A wet stain is air-dried first. 4. Air-dry, then package breathable. This is the rule television always breaks: biological evidence goes into paper — envelopes, bags, or boxes — never sealed plastic (Chapter 2). Sealed plastic traps moisture; moisture breeds bacteria and mold; microbes and humidity degrade DNA. A bloody knife sealed wet in a plastic bag can arrive at the lab with its DNA ruined. 5. Controls and chain of custody. Take substrate controls; label, seal, initial; preserve the unbroken trail (Chapter 2). A perfect profile from a stain with a broken chain of custody may never reach a jury.

The handoff also inherits two interpretive traps from the DNA chapters that bear repeating here, because blood evidence makes both vivid. The first is the transfer problem (Chapter 7, expanded for touch DNA in Chapter 8): finding someone's blood somewhere does not, by itself, establish how or when it got there. Blood can be transferred on a shared towel, tracked on a shoe, carried on a tool, or — rarely but really — relocated by secondary transfer. Presence is not action; presence is not timing. The stain says "this person's blood is here"; it does not say "this person bled here, at the time of the crime, in the act of the crime." Building that bridge requires other evidence — pattern, pathology, sequence — and an analyst who builds it from the blood alone has overreached.

The second trap is mixtures (Chapters 8–9). A bloodstain is not always one person's blood. Two people who bled in a struggle, a victim's blood with a perpetrator's sweat, a stain laid over an older stain — these produce mixed profiles whose interpretation is the hard, contested problem Chapter 9 dissected with the likelihood ratio (LR) and the prosecutor's fallacy. The serologist's job is to recognize that a stain might be mixed and to preserve it so the DNA analyst can sort it out honestly; it is not to assume a single source because that is the tidier story.

⚖️ In the Courtroom The most consequential blood-evidence disasters in American legal history are not chemistry failures — they are collection failures, attacked devastatingly on cross-examination. When the defense in a major case can show that swabs were handled by an analyst who also handled the reference sample, that blood vials were left in a hot truck, that controls were skipped, or that an item was packaged wet, the strength of the eventual DNA profile becomes almost irrelevant. The jury is no longer asked "does the profile match?" but "can you trust where this sample has been?" This is why the chapter spends so long on the unglamorous swab. The ceiling on a DNA result is set at the scene, by the person holding the paper bag. (We study the most famous example, the O.J. Simpson collection record, in this chapter's first case study.)

Theme 1 lives in this handoff too. The cleanest, most defensible thing a blood-plus-DNA workup often produces is an exclusion: the stain's profile does not match a given suspect, full stop, and that person is out. Inclusions, as always, come wrapped in a probability and the transfer caveat. Collect as if the most important result you can produce is the one that clears someone — because, statistically, it frequently is.

10.4 Bloodstain pattern analysis: the physics of spatter

Now we cross from chemistry into physics, and from the strong end of the validity spectrum toward its contested middle. Bloodstain pattern analysis (BPA) is the examination of the size, shape, distribution, and location of bloodstains to infer the events, mechanisms, and sequence that produced them. Where serology asks what and whose, BPA asks what happened — and that ambition is exactly what makes it powerful in theory and perilous in practice. A pattern analyst tries to read, from dried marks on a wall, the violent motion that no one witnessed.

The discipline does rest on some genuine physics, and it is worth taking that physics seriously before we take its limits seriously. Blood is a fluid with a roughly constant surface tension and viscosity, and a drop of blood in flight behaves predictably. Three real, defensible principles anchor the field:

1. A free-falling drop of blood forms a sphere, not a teardrop. Surface tension pulls an airborne drop into a sphere (the "teardrop" of cartoons is a myth). When that sphere strikes a surface, the stain it leaves encodes information about how it traveled.

2. The shape of a stain encodes its direction and angle of impact. A drop striking a surface straight down (90°) leaves a round stain. A drop striking at an angle leaves an elongated stain — an ellipse — and the more acute the angle, the longer and narrower the ellipse, often with a small "tail" or spines pointing in the direction of travel. This is the one quantitative idea in BPA, and it is real geometry: the angle of impact can be estimated from the stain's dimensions, because the ratio of the ellipse's width to its length is the sine of the impact angle.

$$\sin(\text{angle of impact}) = \frac{\text{width of stain}}{\text{length of stain}}$$

A nearly round stain (width ≈ length) gives an angle near 90°; a long, thin stain (width much less than length) gives a shallow angle. This is the most defensible measurement in BPA — a directly measurable geometric relationship — and even it carries error, because the surface texture, the stain's edge irregularity, and the analyst's choice of where the ellipse "ends" all introduce uncertainty.

3. Drop size and distribution correlate (loosely) with the energy of the event. Lower-energy events (blood dripping from a wound, or a hand) tend to produce larger drops; higher-energy events (an impact, a gunshot) tend to atomize blood into many smaller droplets. This is the basis for the old categories — passive stains (drops, flows, pools formed by gravity), transfer/contact stains (a bloody object pressed against a surface, e.g., a hand-print or a wipe), and spatter stains (blood dispersed through the air by a force).

Spatter — projected blood dispersed as droplets by an applied force — is the category that draws all the courtroom attention and all the controversy. Note carefully what "spatter" does and does not mean. It is a physical description of dispersed droplets, not a conclusion about what applied the force. The same general field of small droplets can, in principle, arise from very different events, and distinguishing "an impact struck a wet bloodstain," "an arterial breach," "blood flung off a swinging object (cast-off)," and an "expirated" pattern (blood blown out through the nose or mouth) is exactly where reliable physics shades into interpretive judgment.

   READING A SINGLE SPATTER STAIN (schematic; not to scale)
   ─────────────────────────────────────────────────────────

      drop strikes near 90°        drop strikes at a shallow angle
            (overhead)                    (glancing)

             ●                          ◖━━━━━► tail / spines point
          round                              in the direction of travel
       width ≈ length                    elongated ellipse
       angle ≈ 90°                       width << length  →  small angle
                                         angle ≈ arcsin(width / length)

   DIRECTIONALITY: the tail/narrow end points the way the drop was traveling.
   ANGLE: from the width-to-length ratio of the ellipse (the one real measurement).
   ── These two facts are defensible. "What event made the spatter" is an INFERENCE. ──

🔬 At the Bench Modern BPA still measures directionality and impact angle stain by stain — a tedious process of photographing, scaling, and measuring dozens or hundreds of individual stains in a pattern. Done rigorously, it produces two relatively defensible outputs: the direction each drop was traveling, and the angle at which each struck. Trained analysts also distinguish pattern categories (passive vs. transfer vs. spatter) with reasonable reliability when the patterns are clear. The trouble begins when the question escalates from "what is the geometry of these stains?" to "what human action produced them, in what sequence, by whom?" — because the physics thins out fast as the questions get more specific, and the analyst's expectations rush in to fill the gap (10.6).

🧠 Cognitive-Bias Watch BPA is a textbook host for contextual bias (Theme 3; named fully in Chapter 31). The pattern on the wall is ambiguous; the analyst is frequently told the suspect's account, the victim's injuries, and what the detective believes happened before interpreting the stains. Knowing "the husband says she fell, but we think he beat her" primes an analyst to see "impact spatter from a beating" in a pattern that is also consistent with other mechanisms. Because BPA interpretation is subjective and the "right" answer is not pinned by a measurement, expectation can steer it without the analyst ever noticing. The fix is the same one this whole book keeps prescribing: keep the analyst blind to the desired conclusion (context management; Chapter 31) and document the stains' measurable features before forming an account of events.

10.5 Area of origin and what BPA can support

The most ambitious thing BPA attempts is the area of origin — the estimated location in three-dimensional space from which a spatter pattern originated. If you can determine, for each of many stains, both its direction of travel and its impact angle, then in principle you can project each stain backward along its flight path and find where those paths converge. The intersection in two dimensions (on the floor or wall) is the area of convergence; extended into the third dimension using the impact angles, it becomes the area of origin — an estimate of where in the room the blood source (a wound, a struck surface) was at the instant the spatter was created.

   AREA OF ORIGIN — THE STRINGING / TANGENT METHOD (schematic, not to scale)
   ─────────────────────────────────────────────────────────────────────────

      WALL ▣                       Each stain → a string (or computed line)
       │   \   stain 1 ──┐         laid back along its direction, angled up
       │    \            │         by its impact angle.
       │  string         │
       │  angled up      ▼
       │   \         ╳  ← area of CONVERGENCE (2-D, on the surface)
       │    \       /│\        ↑
       │     \     / │ \       │  lift each string at its impact angle →
       │      \   /  │  \      │  strings cross in space at a fuzzy region:
   floor ─────●─────●───●──────┘     the AREA OF ORIGIN (3-D, a volume — not a point)
            stains on the floor

   The classic physical version uses actual STRINGS taped to each stain;
   modern labs compute the same geometry in software (e.g., trajectory programs).
   THE OUTPUT IS A REGION WITH REAL UNCERTAINTY, never a precise coordinate.

Done carefully, an area-of-origin estimate can support genuinely useful, limited conclusions — and the word "limited" is doing heavy lifting. It can suggest, for instance, that a spatter-producing event occurred low to the floor rather than at standing height — which might be consistent with a victim who was already down when struck, and inconsistent with an account in which the victim was standing. That is a real contribution: not "here is the murder," but "this account is hard to reconcile with this physics, and that one is easier." Note the verbs. Area of origin, at its honest best, excludes or fails to exclude particular accounts of where a person was when blood was projected. It is, like so much else in this book, strongest as an instrument of exclusion (Theme 1).

But every step of the reconstruction injects error, and an honest analyst names each source out loud:

• The straight-line assumption ignores gravity. Projecting a stain "straight back" along its impact direction treats the droplet's path as a straight line, when in reality blood droplets follow a curved, parabolic arc as gravity pulls them down. The straight-line (tangent) method therefore systematically places the origin higher than the true point. Software can partly correct for this, but only with assumptions about droplet velocity that are themselves uncertain.
• Stain selection is a judgment call. An analyst chooses which stains to include; well-formed, clearly elliptical stains near the pattern's edge are favored, but the choice introduces subjectivity, and different analysts may select different stains and reach different origins.
• Surface and measurement error compound. Texture, absorbency, and the analyst's choice of where each ellipse begins and ends all perturb the angle estimate, and small angle errors fan out into large position errors over the distance of the projection.
• It says nothing about sequence, weapon, or actor. An area of origin locates where a spatter event happened in space. It does not say what caused it, in what order relative to other events, with what weapon, or by whom. Those are separate inferences requiring separate evidence.

⚖️ In the Courtroom The defensible BPA testimony sounds modest and specific: "The spatter on the north wall shows directionality consistent with a source low and to the left; the area-of-origin estimate places that source roughly 30 to 50 centimeters above the floor, with substantial uncertainty." The indefensible testimony sounds like a screenplay: "The defendant stood over the victim and struck her four times with a downward motion." The first describes geometry and states its uncertainty; the second narrates a film the blood cannot license. On cross, the killer questions are simple and rarely survivable: What is the error on that origin estimate? Did you account for the parabolic path? What other mechanisms could produce this pattern? Were you told what the detective believed before you interpreted it? An analyst who cannot answer the first three, or who must answer "yes" to the fourth, has handed the defense the case.

🔍 Check Your Understanding 1. Why does the straight-line "stringing" method tend to place the area of origin too high? 2. An analyst concludes the victim was "kneeling when struck." Which part of that is potentially supportable by area-of-origin physics, and which part is narration the blood cannot supply?

10.6 The limits and disputes of BPA (a PCAST-era flagged method)

We now place bloodstain pattern analysis honestly on the validity spectrum (Theme 2), and the placement is not flattering. BPA occupies a genuinely uncomfortable position: it has a real physical core (the geometry of 10.4 and the limited area-of-origin estimates of 10.5) wrapped in a great deal of subjective interpretation that has repeatedly been presented to juries with far more confidence than the science supports. It is one of the clearest cases in the book of a method whose careful, modest version is useful and whose ambitious, narrative version is dangerous — and the two are practiced under the same name, often by the same person, sometimes in the same testimony.

The field's own reckoning has been blunt. The 2009 NAS report singled bloodstain pattern analysis out for unusually pointed criticism, observing that the interpretations are often more subjective than scientific, that BPA training and qualifications vary widely and are sometimes minimal, that the underlying studies establishing reliability were thin, and — memorably — that the uncertainties in BPA conclusions are enormous and analysts' opinions are more subjective than has been acknowledged in court. The report did not say blood patterns are meaningless; it said the discipline had not done the work to justify the confidence with which its conclusions were delivered. In the years since, the broader scientific critique captured by the PCAST-era standard — does a feature-comparison method have measured reliability, with a known error rate, established by properly designed black-box studies? — has reinforced the concern. The honest summary: BPA's directionality and impact-angle measurements are defensible; its event-reconstruction and "this is what the suspect did" narratives sit on weak empirical ground, with error rates that have not been adequately established.

⚠️ Junk-Science Alert The most dangerous BPA testimony is also the most cinematic: an analyst describing, blow by blow, how a killing unfolded — the number of strikes, the relative positions of attacker and victim, the assailant's handedness — read off the spatter "like a book." Some of that may be loosely supportable; much of it is the analyst's imagination wearing the authority of physics. Be especially wary when (a) the conclusion is far more specific than directionality and angle can justify, (b) the analyst's training is brief (some practitioners hold only a short course), and (c) the analyst was steeped in the investigative theory before interpreting the stains. The combination of high specificity, thin qualification, and contextual contamination is a flashing warning light. As with bite marks (Chapter 16) and microscopic hair (Chapter 19), the presence of a confident expert is not evidence of validity — the error rate is, and for event-level BPA conclusions, that error rate is largely unmeasured.

The courts have produced cautionary tales that make the stakes concrete. There are documented cases in which BPA testimony helped convict and was later seriously disputed or contributed to an overturned conviction — most prominently the David Camm case in Indiana, where competing pattern analysts reached opposite conclusions about whether stains on a shirt were high-velocity impact spatter (implying the wearer was present at the shooting) or transfer stains (implying innocent contact with bodies afterward). Camm was tried three times across more than a decade; the bitterly contested, irreconcilable BPA opinions in that litigation became a national emblem of the discipline's subjectivity. The same stains, examined by qualified experts, supported opposite stories. That is the defining symptom of a method that has not pinned its conclusions to measurable reality. (We study Camm closely in this chapter's second case study.) The lesson is not "BPA is fraud." It is "BPA's event-level conclusions are opinions, sometimes deeply contested ones, and a jury must be told so."

⚖️ In the Courtroom What may a careful BPA analyst honestly say? Roughly this much, and not more: identify pattern categories where they are clear (passive, transfer, spatter); state directionality from stain shape; estimate impact angles from stain geometry with stated uncertainty; offer an area-of-origin estimate as a region with explicit error; and say a given account is consistent with or difficult to reconcile with the physical pattern. What the analyst must not do: narrate the crime, assign specific actions to specific people, claim a precise origin point, declare a single mechanism to the exclusion of others without basis, or present any of it "to a reasonable degree of scientific certainty" as if it were measured. The CSI effect (Theme 4) makes this especially hazardous: jurors primed by television expect the blood to "tell the story," and an analyst who obliges them — who gives the confident reconstruction the jury is hungry for — is feeding a fiction the science cannot back.

Where does that leave the working investigator and the careful juror? With a split verdict, which is the honest one. Use BPA's strong parts — document patterns meticulously, measure directionality and angle, estimate area of origin as a bounded region, and let it exclude accounts it can exclude. Distrust BPA's ambitious parts — the blow-by-blow reconstruction, the confident single mechanism, the narrative that conveniently matches the detective's theory. And insist, always, on the safeguard: an analyst who interprets blood patterns while blind to the case theory is doing science; an analyst who interprets them to confirm the case theory is doing something else, no matter how many stains they measured.

🗂️ The Case File

The doorframe stain, and the spatter that didn't fit a collapsing house. Two pieces of blood evidence from the Mill Creek cabin reach the bench this chapter, and together they begin to pull a loose thread in the "accidental fire" story.

The first is the stain introduced in this book's signature figure — the dried reddish-brown mark on the interior doorframe between the front and back rooms, about 40 cm above the floor, its upper edge partially charred. The CSI documented it before touching it, presumptively tested its edge (Kastle-Meyer: pink), and collected it by the wet-then-dry method, air-drying the swab and packaging it in a breathable box. At the state lab, confirmatory testing identified it as human blood, and it is queued for DNA (Chapters 7–9 supply the typing; the result is pursued there, not here). The single most important physical fact this chapter can defend about it is one of sequence: the lower edge of the stain lies below the char line, which means the blood was deposited before the fire reached that height. Blood present before the fire is consistent with an injury preceding the burn — a thread worth pulling, not a conclusion.

text FIGURE 10.3 — "The stain by the doorframe" [the cold case] THE ITEM A dried reddish-brown stain, ~8 cm, on the interior doorframe of the Mill Creek cabin, about 40 cm above the floor, partially charred at its upper edge. THE CONTEXT Collected on day 2, photographed in place with a scale, swabbed (wet-then-dry), the swab air-dried and packaged in a breathable box — not a plastic bag. WHAT IT SHOWS Presumptive testing (Kastle-Meyer) is positive; confirmatory testing identifies it as human blood. Its lower edge is below the char line, so it predates the fire. WHAT IT DOESN'T It does not, by itself, say whose blood it is, when it was deposited, or that a crime occurred — people bleed in houses for innocent reasons. Presumptive tests can false-positive on some plant peroxidases. THE INFERENCE Blood present before the fire is *consistent with* injury preceding the burn — a thread worth pulling, not a conclusion. DNA (Ch. 7) and pattern (this chapter) come next. THE LESSON Sequence is everything: "before the fire" is a defensible physical fact; "the victim was attacked here" is a story the evidence has not yet earned.

The second piece is a small spatter pattern on the same doorframe and the adjacent wall, low to the floor. A BPA analyst measured the clearer stains for directionality and impact angle and noted, cautiously, that the geometry is consistent with a low source — and, more pointedly, that the spatter's character is not what one would expect from blood relocated by a structural collapse during the fire (falling debris, dripping, charring artifacts), but rather with blood projected by a force while it was still liquid, before the fire. The analyst wrote the careful version: directionality and angle measured; area of origin estimated as a low region with substantial uncertainty; the pattern inconsistent with a pure collapse-in-fire explanation and consistent with a pre-fire impact. No blow-by-blow. No named actor.

Honest status after this chapter: the blood evidence shifts the picture without resolving it. Force was likely applied before the fire — both the pre-char doorframe stain and the low spatter point that way — which is hard to square with a simple accident and easy to square with something staged. But this chapter cannot say whose blood it is (DNA, pending), cannot say who applied the force, and cannot, on BPA alone, narrate what happened. We have a defensible sequence (blood, then fire) and a defensible exclusion (the spatter is hard to reconcile with collapse-in-fire). The autopsy (Chapter 11) will test the "before the fire" thread far more decisively. Who is excluded / who remains: unchanged from Chapter 9 — Salas and Dana remain on the path to exclusion pending DNA; Keller remains consistent with the gas-can mixture but unproven; Renner's confession still hangs unexamined. The blood adds sequence, not a suspect.

Conclusion

Blood evidence splits cleanly into two questions, and the entire chapter has been an argument for never confusing them. What is it, and whose is it? is a serology-into-DNA question that sits near the strong end of the validity spectrum: presumptive screens like Kastle-Meyer and luminol triage fast (with named false positives), confirmatory and species tests establish that it is human blood, and DNA — collected by the unglamorous, decisive ritual of the breathable paper bag — answers whose. The error in this half is almost never the chemistry; it is treating a presumptive glow as an identification, or hanging a narrative on it, or destroying the sample at collection.

What happened? is a bloodstain-pattern question, and it sits in the contested middle. BPA has a real physical core — directionality from stain shape, impact angle from the width-to-length ratio, area of origin as a bounded region — and those modest measurements can honestly exclude accounts of where a person was when blood was projected. But its ambitious, cinematic version — the blow-by-blow reconstruction, the single confident mechanism, the story that happens to match the detective's theory — rests on subjective judgment that the 2009 NAS report and the PCAST-era standard found unvalidated, with error rates that remain largely unmeasured. The David Camm case is the lesson in one sentence: the same stains, read by qualified experts, supported opposite stories.

In the cold case, the blood has given us sequence — deposited before the fire — and a spatter pattern hard to reconcile with a collapsing, burning house. That is a thread, not a verdict. In the next chapter, the autopsy takes that thread and pulls hard: forensic pathology will ask whether Marcus Diallo was alive when the fire started at all, and the answer will do what no bloodstain alone can — overturn the premise of the entire case.

Key Terms

• Serology — the forensic identification, characterization, and (historically) grouping of blood and other body fluids; today focused on is a fluid present, and which one.
• Presumptive test — a fast, sensitive, inexpensive screening test indicating a substance is probably present (and reliably indicating where it is not); never an identification by itself, because it can false-positive.
• Confirmatory test — a more specific test that establishes, to high certainty, that a substance is the target (e.g., that a stain is human blood).
• Kastle-Meyer — the classic presumptive blood test using phenolphthalein and hydrogen peroxide, turning pink in the presence of blood's heme/peroxidase activity; sensitive but not specific.
• Bloodstain pattern analysis (BPA) — the examination of the size, shape, distribution, and location of bloodstains to infer the events and mechanisms that produced them; physically grounded in part, but heavily interpretive and flagged by the NAS 2009 report.
• Spatter — blood dispersed through the air as droplets by an applied force; a physical description of dispersed drops, not a conclusion about what force caused them.
• Area of origin — the estimated three-dimensional region from which a spatter pattern originated, reconstructed from stains' directionality and impact angles; a bounded estimate with real uncertainty, not a precise point.
• Luminol — a presumptive blood reagent that chemiluminesces (glows blue in darkness) with heme; reveals latent/cleaned-up bloodstains but false-positives on bleach, metal ions, and plant peroxidases, and can degrade DNA.

Spaced Review

1. A bloodstain on a doorframe is identified as human blood, and DNA later matches a suspect. Why does that still not establish that the suspect bled there during the crime? Name the interpretive trap, and the chapter that owns it. (§10.3; Chapters 7–8.)
2. The gas-can stain in the cold case was a mixture whose interpretation required a likelihood ratio (Chapter 9). Restate the prosecutor's fallacy in one sentence, and say why a serologist's job is to preserve a possibly-mixed stain rather than assume a single source. (§10.3; Chapter 9.)
3. From Chapter 3: a bloody knife is sealed wet in a plastic bag and sent to the lab. Which property of physical evidence has been violated, and what is the likely consequence for the DNA? (§10.3; Chapter 3, evidence integrity.)
4. From Chapter 1: classify each as class or individual information — (a) "this stain is human blood," (b) "this stain's DNA profile matches one person with a random match probability of 1 in a billion." Which is closer to individualization, and why? (§10.3; Chapter 1.)
5. Validity-spectrum question: Where does bloodstain pattern analysis sit on the NAS/PCAST spectrum, and why does it sit there — distinguishing the part of BPA that is defensible from the part the NAS 2009 report criticized? (§10.4–10.6.)