7.13 Temporal Sequence Ambiguity

Temporal sequence ambiguity refers to the uncertainty about the true causal ordering of events that arises when variables are coupled through circular causal relationships operating across different time scales, when the temporal resolution of observation is insufficient to resolve the actual sequence of cause and effect, or when the feedback delays in a circular system make the identification of earlier versus later causes theoretically or practically impossible. In circularly causal systems, events that appear simultaneous at one level of temporal resolution may have a definite causal ordering at a finer resolution; conversely, events with clear temporal succession may nevertheless produce causal ambiguity when the faster feedback pathways can cycle effects back to earlier variables before the observer's measurement captures the intermediate states.

The most fundamental source of temporal sequence ambiguity in circularly causal systems is the relationship between the observation timescale and the feedback delay. In a circular system in which variable A influences variable B with delay τ₁ and variable B influences variable A with delay τ₂, observations made at intervals much longer than τ₁ + τ₂ cannot resolve the causal sequence: each observation sees only the joint state of A and B, with no information about which changed first during the inter-observation interval. The Granger causality framework addresses this by asking whether past values of A help predict current values of B, net of past values of B itself:

If the coefficients β_k are jointly significant, A is said to Granger-cause B. But in a truly circular system where B also Granger-causes A, both directions of Granger causality will be detected, and no unambiguous temporal ordering can be established from the data alone—the temporal sequence ambiguity persists even with high-frequency data and statistical analysis.

Contemporaneous causation—when two variables influence each other within the same observation interval—produces the most severe form of temporal sequence ambiguity. In macroeconomic systems, consumption and income influence each other within the same quarter: higher income leads to higher consumption in the current quarter, but higher consumer spending also generates higher income in the current quarter through the demand multiplier. Quarterly data cannot resolve which changed first, and the contemporaneous mutual causation produces a simultaneous equations problem that requires additional identifying assumptions (instrumental variables, structural model constraints) to estimate causal effects.

Temporal sequence ambiguity in biological systems arises when multiple feedback pathways operate at very different timescales, creating situations where the same event can be characterized as both cause and effect depending on which feedback pathway is being analyzed. In the regulation of blood pressure, the baroreceptor reflex produces compensatory responses within seconds of a pressure change, while hormonal regulation through the renin-angiotensin-aldosterone system adjusts plasma volume over hours to days, and structural vascular remodeling adapts over weeks to months. A sustained increase in blood pressure will engage all three systems, each feeding back on pressure through a different pathway and timescale. Asking which factor "caused" the eventual new pressure setpoint is temporally ambiguous because the answer differs depending on which feedback pathway one considers and which observation timescale one adopts.

In econometric analysis, temporal sequence ambiguity motivates the use of structural vector autoregressive (SVAR) models that impose theoretical causal orderings on contemporaneously correlated variables. The Cholesky decomposition commonly used to identify SVAR models assumes that the variable ordered first has no contemporaneous effect from the variables ordered later, while the variable ordered last can be contemporaneously affected by all other variables. This identification assumption resolves the temporal sequence ambiguity by imposing a causal hierarchy based on prior economic theory, but the resolution is theory-dependent: different orderings produce different identified causal effects, and the ambiguity about causal ordering cannot be resolved from the data alone without additional structural assumptions.

In historical analysis, temporal sequence ambiguity challenges explanations of causation in complex social processes. When an economic recession and political instability co-occur, and both are plausibly causes and effects of each other through multiple feedback pathways, asserting a clean temporal ordering of cause and effect misrepresents the circular causal structure that generated the observed dynamics. The historical record—which documents events in chronological order—creates an appearance of temporal sequence that can be mistaken for evidence of causal priority, when the actual dynamics were circularly causal with both phenomena reinforcing each other simultaneously through different feedback pathways operating on different timescales. Robust historical causal analysis requires examining the full structure of the feedback mechanisms in operation, not merely the sequence in which events were chronicled.