What does it mean to understand a phenomenon scientifically?

• It should be possible to provide a concise description of many phenomena in terms of general, abstract concepts or principles
  • Newton's laws of motion are an excellent example of this. The equation $F = m a$, i.e., force is equal to mass times acceleration, summarizes many different phenomena succinctly. It also provides a guide to scientists and engineers who want to develop models of physical systems.

• A range of phenomena can be explained in terms of underlying mechanisms:
  • Statistical mechanics as the basis for thermodynamics.
  • Molecular biology as the basis for cell function.
  • Cell function as the basis for organ function.
  • Organ functions as the basis for understanding whole organisms.
  • How organisms interact as the basis for understanding ecosystems.

• The general scientific process: Description, Prediction, Control:
  • Description: A wide range of phenomena are summarized, usually through mathematical equations, and related to underlying mechanisms.
  • Prediction: Based on these descriptions, we can predict how the system will behave in the future, and in response to perturbations.
  • Control: Once we can describe and predict phenomena, we can often develop ways of controlling them.

• This process is the reason that basic research, medicine, engineering, and technology are so intertwined.

• A successful example: Mechanics
  • Newton's laws provided the basis for the initial description; these laws needed to be modified for very small, very fast, or very massive objects (i.e., those to which quantum mechanics, special relativity, or general relativity apply).
  • Based on these laws, especially for macroscopic objects moving at speeds that are small relative to the speed of light, it is often possible to predict their trajectories for extended times.
  • In turn, this makes it possible to predict the trajectories of projectiles and rockets, though adjustments may need to be made during movement.

A moderately successful example: Weather prediction

• • Description:
    • Equations: Navier Stokes equations (fluid flow) in a rotating inertial frame, continuity equations (conservation of mass), thermal equations (temperature as a function of heat sources and sinks)
    • Data: Sensors at multiple temporal and spatial scales (e.g., Doppler radar)
    • Multiscale modeling: localities, regions, countries, the entire planet
    • Conceptual frameworks: warm fronts, cold fronts

• • Prediction:
    • High performance computer clusters, ensembles of models (different initial conditions, different parameter values)
    • Predicting tracks of hurricanes or tornadoes
    • Obstacles: deterministic chaos, turbulence

• • Control:
    • Little progress based on local control
    • Inadvertent effect of releasing large amounts of carbon dioxide and other greenhouse gasses on the entire planet

• What would a successful neuroscience look like?