Surprise Is The Logarithm Of Reality

Surprise is the mathematical distance between Reality and Expectation.

Reality is the quotient of the Actual-Expectation relation:

R(t) = A(t) / E(t)

But Reality by itself is not yet surprise.

Surprise is the natural logarithmic function of Reality.

Because Reality is complex, the formal model uses the complex logarithm:

S(t) = Log(R(t))

That is the next step in the ladder.

The Shape Of Complex Surprise

Complex surprise can be written:

S(t) = ln|R(t)| + i Arg(R(t))

Or:

S(t) = X(t) + iY(t)

where:

X(t) = ln|R(t)|
Y(t) = Arg(R(t))

The real component measures scalar surprise.

The imaginary component carries angular or orientational surprise.

In ordinary language: surprise is not only how much Reality differs from Expectation. It also contains orientation.

The Simple Scalar Case

For public explanation, the scalar case is useful.

If R is real and positive:

S = ln(R)

Then:

R = 1 means S = 0
R > 1 means S > 0
R < 1 means S < 0

In this simplified scalar case:

D = |ln(R)|

This is attention-demand only in the simplified scalar picture.

The full model is larger.

Why The Scalar Case Is Incomplete

The scalar case explains one intuition:

The farther Reality is from Expectation, the more surprise appears.

But human attention is not one instantaneous scalar surprise value.

The full model requires three corrections:

1. Reality is complex because Expectation is complex.
2. Surprise is Log(R) when Reality is complex.
3. Conscious attention appears only after surprise is accumulated and renormalized across an attention window.

So the formal ladder is not:

Reality -> attention

It is:

Reality -> complex surprise -> accumulated surprise -> renormalized attention pointer

A Note On Branches

The complex logarithm is multi-valued:

Log(R) = ln|R| + i(arg(R) + 2*pi*k)

For a point estimate, use the principal branch.

For a time-series model, unwrap the phase so that artificial jumps at -pi and pi do not create false attention spikes.

This is a technical detail, but it matters. A formal model of attention cannot allow notation artifacts to masquerade as surprise.

The Key Point

Surprise is not merely shock.

Surprise is the mathematical difference between Reality and Expectation, expressed through:

S(t) = Log(R(t))

But even surprise is not yet attention.

Next: Attention is normalized accumulated surprise.