Where does contamination collide?

The turning of an object into something undesirable or dangerous through contact with a source seen as tainted, lasting even after any physical trace is gone. Frazer's 1890 The Golden Bough named the law of contagion, the belief that things once in contact keep acting on each other, as one of the two principles of sympathetic magic. Rozin, Millman, and Nemeroff's 1986 study brought this idea into experimental psychology as contamination, showing that many people refuse juice that a sterilised cockroach has briefly touched or a laundered shirt that a disliked person once wore. Here the referent is a judgment in the mind about a tainted object, not microbes multiplying in a dish, a mixed manuscript, or a leak between study groups.[1][2]

Unit

One judged instance of an object made undesirable by contact

Recovery means

Not applicable

Measured by

Assessed by willingness-to-contact ratings and behavioural disgust tasks in the laboratory

A condition of a copied text whose readings descend from more than one source rather than a single line of copying, so it cannot be placed on a simple family tree of witnesses. The label comes from the Latin contaminatio, used in antiquity for combining separate works into one, and it entered manuscript study in the nineteenth century as editors traced how scribes mixed exemplars. Maas's 1958 Textual Criticism gave the standard account, warning that there is no sure remedy against it because a blended witness hides its own line of descent. Here the referent is a pattern of borrowing among written copies, not microbes in a medium, a belief about a tainted object, or a leak between study groups.[3]

Unit

One reading in a witness traced to a second source line

Recovery means

Not applicable

Measured by

Inferred from conflicting shared readings that break a single-descent stemma

The unintended spread of a treatment to people who were meant to serve as the untreated comparison, so the gap between groups shrinks and the effect of the treatment is understated. Cook and Campbell's 1979 Quasi-Experimentation named diffusion or imitation of treatments as a threat to a study's internal validity, describing how control participants learn of and adopt what the treated group receives. Shadish, Cook, and Campbell's 2002 Experimental and Quasi-Experimental Designs carried the account forward as the standard reference for field research. Here the referent is a leak between conditions in a study, not microbes in a medium, a belief about a tainted object, or a mixed manuscript.[4][5]

Unit

One control participant exposed to the treatment condition

Recovery means

Reduced by separating sites, isolating conditions, or blinding, and rarely undone after the fact

Measured by

Estimated from manipulation checks and reports of cross-condition exposure