Stratigraphic Principles and Harris Matrix

The principles of stratigraphy originated in geology, codified by William Smith and Charles Lyell in the early nineteenth century, but Edward Harris adapted them for urban archaeological sites in the 1970s and they translate directly to forensic casework. There are three, and each is doing real work at every grave site.

1. Law of superposition: in any vertical sequence of undisturbed deposits, the youngest is at the top and the oldest is at the bottom. The grave fill is therefore younger than the natural subsoil it sits within, which is obvious, but the law applies at every scale including the different layers within the fill itself.
2. Law of original horizontality: sediment is deposited roughly horizontally. A steeply inclined layer in the middle of a sequence is evidence of disturbance. A grave fill often displays characteristic tipping lines from the angle at which soil was thrown back in, and these are readable in section.
3. Law of lateral continuity: a deposit originally extended in all horizontal directions until it met a boundary or thinned to nothing. Where it is now absent, something removed it. A grave cut breaks the continuity of the surrounding deposits in a characteristic way that experienced excavators recognise immediately.

A fourth principle is sometimes added for forensic contexts: the law of stratigraphic succession, which holds that each context in a sequence occupied a unique position that can be read by its relationships with adjacent contexts. Harris formalised this as the basis for his Matrix. Together these four ideas mean that a careful excavator can place every context in a sequence even when the physical layers have been disturbed, mixed, or partly destroyed.

A grave is, in stratigraphic terms, an intrusive cut filled with a series of deposits. Understanding what that looks like in section is the first skill a forensic archaeologist needs in the field. The cut itself is usually visible as a vertical or near-vertical face where one earth type meets another in a way that does not match the surrounding horizontal sequence. The fill inside the cut is characteristically disturbed: the soil horizons are broken, mixed, and sometimes inverted.

The fill sequence is rarely simple. In a freshly dug and immediately refilled grave, the primary fill around and above the body often contains mixed subsoil thrown in quickly, sometimes with the topsoil horizon inverted because the first spit out of the hole ends up at the bottom of the backfill. Above that sits a secondary fill of more uniform texture as later material washed or slumped in. In long-established graves, roots and animal activity add further complexity, but the core sequence is usually recoverable with careful trowelling.

Edward Harris published his matrix system in the journal World Archaeology in 1975 and expanded it in his 1979 book Principles of Archaeological Stratigraphy. The system was developed for urban sites where hundreds of overlapping cuts and fills create a sequence too complex to hold in memory. For forensic cases, sites are simpler, but the formal discipline of the Matrix is just as valuable because it creates a product that can be challenged and defended in court.

1. Number every context
   Each cut, fill, and natural deposit receives a unique number as it is identified during excavation. Numbers are assigned sequentially in order of discovery, not in stratigraphic order. This means context 001 may be younger than context 015 if the topsoil was stripped first.
2. Record relationships
   For each context, the excavator records which contexts it is above (i.e., younger than), which it is below (older than), and which it is contemporary with (same deposit, different area). These go on the context recording sheet before the context is removed.
3. Draw the initial matrix
   Each context becomes a box. Lines connect contexts according to their recorded relationships: a line from box A to box B means A is older than B. Later material sits higher on the page. The matrix is drawn from field records, not from memory, so it is independently verifiable.
4. Check for redundant links and loops
   A redundant link (A older than B, B older than C, and also A directly linked to C) is removed; the indirect relationship already exists. A loop (A older than B older than A) signals a recording error and must be resolved by re-checking the field notes.
5. Phase the matrix
   Groups of contexts that together represent a single episode of activity are grouped into phases. In a grave context, Phase 1 might be the digging and primary burial, Phase 2 the backfill, and Phase 3 any later disturbance such as a second interment or mechanical damage.

One of the most practically important skills in forensic excavation is distinguishing deposits that formed through natural processes from those created by human action. The consequences of getting it wrong are significant: calling a natural feature a grave shaft means wasting investigation resources and potentially presenting false evidence in court. Calling a grave a natural feature means a body may never be correctly located.

Natural processes that create cut-like features include tree-throw (a windblown tree's root plate pulling up a void that fills with mixed soil), frost cracking and ice-wedge polygons in northern climates, animal burrows and badger sets, and solutional pipes in chalk or limestone where acidic water dissolves a vertical channel. Each of these has characteristic signatures. Tree-throw pits are typically ovate in plan with a characteristic hummock beside them. Animal burrows have smooth, rounded profiles. Human-made grave cuts tend to have sharper, more regular profiles, straight or slightly incurving sides, and a flat or slightly rounded base.

• Plough disturbance: repeated ploughing creates a consistent mixed layer (ploughsoil) with a sharp basal horizon called the plough scar. Graves cut through this horizon will be visible below it if the upper part has not been destroyed.
• Pedogenesis: natural soil-forming processes create the A, B, and C horizon sequence. Interruption of this sequence by a vertical or sub-vertical boundary is the first indicator of an intrusive feature.
• Root channels: large root systems can create sinuous discoloured channels that run through the section. These are usually narrower and more irregular in profile than a grave cut and trace back to the surface root system of a specific tree.

Soil does not form overnight. In temperate climates, developing a recognisable topsoil horizon over a disturbed surface takes decades, and developing a fully differentiated A-B-C soil profile takes centuries to millennia. This makes pedogenesis a relative dating tool. If a grave fill has begun to develop a thin topsoil over it, the burial predates that soil formation. If it has not, the burial is likely recent.

More practically, the degree to which a disturbed fill has re-pedogenised since the burial can help estimate burial interval. Soil micromorphology, the analysis of thin sections of undisturbed sediment under a petrographic microscope, can show whether post-burial biological activity (earthworm calcite granules, fungal hyphae, root channels) has penetrated the fill to a depth that would take a known number of years at measured growth rates. This is specialist work but it is increasingly available for high-value cases.

Ploughing horizons add another dating layer. A grave whose upper fill has been truncated by ploughing and whose surviving lower portion sits below the plough scar must have been dug before ploughing began at that depth. If the field was last deeply ploughed in a documented year, the burial is at least that old. The converse is also useful: if the grave cut is entirely within the ploughsoil, it may postdate the last period of ploughing.

In a forensic case, the stratigraphic sequence is not just an academic exercise. It directly supports or undermines specific investigative hypotheses. If a suspect claims they only visited the site once, a stratigraphic sequence that shows two separate digging events contradicts that account. If a defence argues that remains were moved to the site after death from somewhere else, the presence of a well-defined grave cut filled with material consistent with a single digging event supports a primary interment.

The power of the sequence is that it survives even when other evidence degrades. DNA may be unrecoverable from an old burial. Clothing may have decomposed. But the stratigraphic relationships between context 007 (the grave cut) and context 004 (the natural subsoil it intrudes into) are physically preserved in the ground and will remain consistent no matter when the excavation takes place. An investigator who documents those relationships correctly in 2024 creates a record that a court could review in 2034.