WHAT IS A BLACKBOX?
|------------------| Known Input| |Known Output ---------->| Unknown Interior |------------> |(mechanics? math?)| |__________________|
As shown, you KNOW the INPUT and OUTPUT of the blackbox, but know nothing about what happens inside the black box to transform INPUT to OUTPUT. This is covered by speculation:

Some, such as the famous British physicist, Lord Kelvin (x-y), inferred mechanical devices in the blackbox transforming INPUT to OUTPUT;
some, such as his friend and fellow Scotsman, William Clerk Maxwell (x-y), write mathematical equations to "connect" INPUT to OUTPUT.
Maxwell's famous equations of electricity and magnetism did much moch more:

united electricity and magnetism and optics;
articulated a spectrum of oscillations of the electromagnetic field;
this spectrum included not only the light spectrum bu predicted radio waves, detected after Maxwell's death by Heinrich Hertz (x-y).
When Kelvin first heard about Hertz's work, he denouced it as a hoax, which intimidated most physicists from investigating it, until Italian physicist, Guillermo Marconi (x-y) broadcast across the Atlantic. To the end of his life, Kelvin said he didn't understand how radio works -- apparently because he wasn't provided with mechanical devices to activate it.
At some level, scientists must always resort to a "blackbox".
In physical science, the atom was once a blackbox to explain chemical reactions. Progress in "quick & penetrating" photography provided observables which some accept as atoms, placing them outside the box. Atomic constituents, such as electrons, protons, neutrons were blackboxes, but collider photos show events interpreted as "bouncers" or "explosions" of these, perhaps placing them outside the box. But quark contituents of poorotns, neutrons, mesons, are at some perspective still blackbox.
Electrodynamics was "rid" of "infinities" by a blackbox filled with Feynman diagrams of prolonged interaction.
Biology once had a blackbox of cells. Then of cellular constituents. It continues to have a blackbox of evolutionary events of the past -- and similar problems confront geology.
C. S. Peirce (whose semiotics is invoked in a file at this website) taught us about three ternms critical to expistemology ("study of what we know"):

concretions: directly observable (such as a particular tree);
abstractions: formulated classes of concretions (such as tree) -- always reducible to concretions;
illations: unobservable but inferable.
Because of the need for illations (such as past events or experiences of other persons), blackboxes will always be needed. In legal terms, science mostly depends upon "circumstantial evidence" and "hearsay witness".
Bertrand Russell said, "Whenever possible, logical constructs are to be substituted for inferred entities." The "goodies" in blackbox are inferred entities. The best logical constructs are mathematical equations. When they link INPUT to OUTPUT, we are satisfied.