What Do We Actually Observe?

Intuition

We never receive the event itself

When we say that we are looking into the past, several physical and interpretive steps have been compressed into one sentence. The distant event does not travel to us. Electromagnetic radiation propagates, reaches a detector, becomes a local material response, passes through processing, and is finally presented as an image, spectrum, catalogue entry, or fitted parameter.

A detector can measure received flux, observed frequencies, direction, polarisation, timing, and uncertainty. It does not directly read the source’s absolute age, original distance, or complete emitted waveform. Those are reconstructed through a physical model.

This distinction does not make reconstruction arbitrary. It makes the causal chain visible.

The Argument

1. What is directly observed?

A spectrograph provides a received quantity such as F_obs(λ_obs,t_obs) together with calibration, noise, angular information, and detector response. These are properties of the received and processed signal.

2. What is inferred?

An identified feature is compared with a laboratory transition:

From z, a cosmological model may infer luminosity distance, angular-diameter distance, lookback time, intrinsic luminosity, and physical size. The chain is therefore:

3. Where interpretation enters

Interpretation enters through line identification, a model of emission, a model of propagation, detector calibration, and the geometric relation used to convert redshift into distance and age.

4. The alternative question

Between the Beats asks whether part of the inferred cosmic history may instead belong to the history of the signal. It does not deny the measurements. It keeps the propagation operator visible in the inverse problem.

5. What is and is not claimed

The claim is that astronomical source history is an inverse reconstruction. It is not claimed that standard cosmology is false, that redshift is unreal, or that distant sources cannot be reliably identified.

Deep Notes

Forward and inverse problems

A forward problem begins with a source state S and predicts an observation O through a propagation-and-detection operator:

The inverse problem attempts to recover S from O. It is stable only when the operator is known, the data retain enough independent structure, and distinct source histories do not collapse onto the same observation.

Model dependence is not arbitrariness

A model can be tightly constrained, predictive, and supported by independent evidence. The epistemic distinction remains:

Non-unique reconstruction

If two pairs of source and path satisfy 𝓕(S₁,P₁) ≈ 𝓕(S₂,P₂), the observation alone does not uniquely determine the source. The later soft-horizon topic asks when this ambiguity becomes dominant.