Eidetic variation
from 1913 Husserl; developed most fully in Experience and Judgment
Change the thing in imagination, feature by feature, until it stops being the thing — whatever you could not change is what it essentially is.
- Active
- from 1913
- Where
- Gottingen and Freiburg
- Also
- eidetische Variation
- Status
- In use
- Tradition
- Continental procedures
What it is
Eidetic variation looks for what cannot be varied. Take an instance, alter it in imagination in every way you can, and watch for the point at which it ceases to be an instance of the kind at all. Whatever survives every variation is the essence.
It is close to conceptual analysis and reached by a different route. Analysis looks for conditions and tests them against cases; this generates the cases by systematic alteration and reads the conditions off what breaks.
How it works
Take a clear instance. Vary one feature at a time, as freely as imagination allows, and ask each time whether it is still an instance. Note every feature that can go. Note the point where it stops being the thing. What remains through every successful variation is the invariant.
The variations must be free rather than realistic. Restricting yourself to variants that could actually occur imports facts about the world into a question about the kind.
The moves, in order
- Take a clear instance of the kind
- Vary one feature in imagination
- Ask whether it is still an instance
- Repeat across every feature you can alter
- What survives all successful variation is the invariant
The procedure
Proposition For the purposes of this consent, a shed is defined by enclosure and not by use.
- Take the clear case: a four-bay machinery shed, steel frame, roof, three walls, earth floor, used for tractors.
- Vary the use: fill it with hay instead. Still a shed. Fill it with a workshop bench. Still a shed. So use is not invariant.
- Vary the material: build the same structure in timber. Still a shed. Vary the floor: pour concrete. Still a shed.
- Vary the walls: remove them all, leaving roof on posts. This stops being a shed and becomes a carport or a hay stand — the panel divides, which locates the boundary.
- Vary the roof: remove it, leaving three walls. Everybody agrees this is not a shed at all.
- Across every variation the roof survives, and partial enclosure survives with the boundary at some walls rather than none.
Therefore What makes it a shed is a roof and at least partial enclosure, not what is kept in it — which means the consent condition restricting use cannot be enforced by arguing that the building has stopped being a shed.
The variations that failed are as informative as the ones that held. The disagreement about the walls is not a defect in the method; it is the method finding a genuinely fuzzy edge and showing exactly where it is.
What it is good for
Working out what a category is really tracking, particularly when a definition has to be applied to cases nobody anticipated. It is quicker than analysis and generates the awkward cases automatically.
Where it fails
The result is bounded by the imagination of whoever runs it. An invariant is only whatever this person failed to vary away, and someone with a wider range of cases in mind will find fewer invariants. There is no procedure for knowing when the variations have been exhaustive.
It also cannot distinguish an essence from a deeply held habit of classification. If everyone in a culture would stop calling it a shed at the same point, the method reports an invariant where what it has found is a convention — and phenomenology has no way to tell those apart.
Common misconceptions
That the variations must be possible in practice. Husserl is emphatic that they are free: the point is to test the concept, and restricting to realistic variants tests the world instead.
Where it came from
Husserl throughout, most systematically in Experience and Judgment, as the procedure by which phenomenology moves from a particular experience to a claim about a kind.
What became of it
Carried into empirical qualitative research, where it is a standard step in phenomenological analysis, and into design practice under other names.
Lasting influence
Its clearest descendant outside philosophy is the systematic variation of requirements in design and specification work.
Why it still matters
Quietly useful; a good tool for anyone who has to write or apply a definition.
Put another way
Bending a thing to find out where it breaks, without breaking anything.
Texts
Edmund Husserl, Experience and Judgment
Sections 87 to 90 give the fullest account of the procedure and its intended force.
Where to start
Run it yourself on a category you have to apply at work. Twenty minutes will find the edge.
The surprising bit
A method invented to establish necessary essences is most useful in practice for finding out exactly where a category becomes genuinely fuzzy.
In their own words
We let ourselves be guided by the free variation of an example, and read off the invariant
Who held it
Alfred Schutz
Connections
- Relies on
- The epoche, The thought experiment
- Parallel to
- Conceptual analysis
The verdict
A fast and honest way to map a concept's edges, and no basis at all for claims about essences.