Preliminary. I use the term 'object' strictly as belonging in the range of an abstraction process. We have "objects" as responses. I use the phrase [putative "thing"] to refer back to that which we infer comes before our lowest level of abstraction - represented by the event level - "putative" because it means we "project" structure with our bains, and "thing" in quotes to further remind myself and any readers that we can not "know" such putative "things"; we only "know" them in virtue of our experiencing our own objects. Because of our nervous systems, I view objects as dynamic, but with some relative invariance, especially when tagged with "words". I agree with Milton that we do the structuring, but I also allow for relative invariance that extends for longer times where we may properly apply the noun form of the term.
Your opening statement proposed moving from epistemology, which general semantics is/is about, to metaphysics, which general semantics is not about. We use metaphysics to construct models based on definitions and a priori starting points. We use epistemology to study how we might "know", and its softer variant "believe".
Characteristics at every level of abstraction are primary (simple) "structures", but structures (general) combine in relations ordered in various ways to form greater structures in what we perceive through our hierarchically organized brains as hierarchically ordered by complexity.
In this perspective, the simplest structure has no subordinate structures that we perceive. We conceive of it as an "atomic" (indivisible) structure.
In abstraction (a relation going from a domain set to a target set) each structure (consisting of one or more characteristics) maps upward to a structure at the target level, and the target structure may be less or more complex.
You would postulate a domain structure that maps to no target during abstraction. We have lots of those. In the structural differential they are represented by unterminated strings.
Though not a principle of general semantics, evolutionary epistemology would allow for our creation of structures in the target domain which do not come from a sensory abstraction process; they come from the combining of different target structuren in new and unusual ways. The metaphor of re-combinate DNA applies here. Each fragmentary structure in our brain - "representing" (abstracted from) part of a domain structure may combine with another to from a larger structure. In doing so, it may form an "idea" that is creatively new and usefu on occasion, but more often than not, just like evolution, the combination is useless, and fails to be retained. The useful ones are kept. The combination takes place through adaptive resonance where the stimulation of a part of a pattern activates the whole. In this case a "common" sub-structure instantiated in cellular circuits simultaneously activates what was heretofore two separate but overlapping circuits. The result is a "new idea" - the creation at a higher level of abstraction of a (new) "structure" in the brain.
Through time-binding we have a lot of "knowledge" (theory), much of it corroborated by scientific research, about how the brain works that we need to bring to bear on the question.
If we move to metaphysics, we are intensionally building models that are related to our starting premises, and that is what you proposed. We can talk about such languguage all day without getting into the principle idea of general semantics - using scientific methods in our daily lives. We must eventually "filter" the metaphysical structures to deal with only those that have hypothetically empirically testable consequences, so we can go through the process of corroborating the metaphysical model we came up with when we took the observer out of the picture by substituting "structure" for "object". Then we get back to general semantics as modern (using as much of today's "knowledge" as possible) applied (empirically, scientifically tested) open (ready to be change) epistemology (how we know, that is, our abstracting process) using the current scientific and mathematical models of valid reasoning only.
Secondly, Thomas, I'm surprised at you.
You began with IF and followed it with a known to be false hypothesis or condition; surely you know that a conditional IF ... THEN ... statement with a false IF part is always true? It matters not what is in the THEN part. Every body can put anything, true or false in the THEN part but the logic is valid and the overall IF ... THEN ... statement has a truth value of T or TRUE.
For this reason, when I was just a wee lad first learning the truth tables of simply logical connectives, I understood that "going back" and saying IF (false) THEN (anything) is always true.
What do people do? They put something nice that they want in the THEN part, and then proceed to go off in a ever diverging fantasy. It might make them feel good to re-read this fiction, but it's all based on a false start. In logic and mathematics is is known that if you start with a false hypothesis, you can "prove" anything.
Sorry but I consider this an example of "un-sanes" reasoning. Note, this is NOT the same as hypothesizing the future in which the truth value of the hypothesis is NOT known to be false, there we are simply establishing consequential relations that may allow us to, through empirical testing, discover whether the hypothesis is false or merely corroborated by specific events.
For that I like to say
IF wishes were horses THEN all people would ride,
IF wishes were horses THEN we would all be armpit deep in horse manure.
Anybody can pick out anything positive or negative to put in the conditional with a false hypothesis.
So I have practices for many decades - more than I remember - never to apply IF (false about the past) THEN (anything), and I quite often challenge anyone I hear making such a statement.
Now, I know that's not quite what you had in mind by your first statement, but your second statement fits the bill for the simple reason that if we take the observer out of the picture then we can have no "knowledge". Knowledge is largely an attitude held by an observer, most often about some putative "thing" represented by his or her "object".
I presume you are (or will become) familiar with the "brains in a Vat" philosophical discussion.
In your second post you say "our theory tells us..." Theories are conditionally corroborated. The theory does not technically "tell us"; we interprete the theory and give it greater or lesser credibility, but we must not make the mistake of many past generations of "scientists" and others who identified their theories as "true". We only get "true" theories in totally intensional disciplines, such as mathematics, where the truth value is wholly dependent upon the axioms. In empirical science, the "truth value" is still conditional on continued testing until the end of time. Time and time again long accepted "theories" which most held to be the absolute truth have fallen to the uncorroborated category by the mean of prediction failures.
That is one of the major points of general semantics - to hold our theories conditionally on future testing. It's the time N+1 is not times 1, 2, ..., N. principle. Use the thory, but be prepared for it to be wrong.
While evolution is strongly corroborated, it is still a theory, and for me a major paradigm, my current paradigm for epistemology.
Everyone agreeing to something does not make it "true". We have lots of culturally accepted "beliefs" that people act like and take to be "absolutely true", and part of the result of such attitudes is all the global strife being purpetrated in the name of various religions.
You wrote It's very difficult to say to someone that "the earth" as we know it did not exist billions of years ago since there were no observers.
This is a fallacy - the fallacy of confusing metaphysics - what "is" - and epistemology - how we know.
Metaphysics is about possible structures.
Epistemology is about how we know, and that includes about possible structures.
Epistemology incudes the postulate that there exists entities (structures) with the capacity to know (and more weakly to find out about and to believe).
You could say, from an axiomatic view point, that epistemology adds an independent postulate to the metaphysical system.
We can talk about what can be or what can not be all day long, but the prospect of someone knowing about what can be or what can not be raises the question from metaphysics to epistemology with the addition of both the observer and the "how" question.
What was relations among structures becomes relations between this and a particular structure with representational abilities.
A Consider universe with some structures.
B some structures are capable of "knowing"
C If there are no structures capable of knowing then there are no structures.
Do you see that the logic is not valid?
For C to be true, all structures must be capable of knowing. Then, and only then, would it be the case that no observers entails no structure.
Thomas's position is clearly that of realism, which holds that what we experience reflects ("is" in the case of naive realism) structures that predate our own existence. We reason by analogy seeing that people are born and die, therefore what I eperience represents some putative "things" that preceeded my consciousness. As far as we know, this is corroborated by anyone we talk to, except for an American native tribe which I vaguely remember somewhere reading, whose philosopy is stated "When I die, the world ends.".
(See also the science fiction short story "All You Zombies". for a fun treatment of this topic.)
People believe its a "law of nature" based on mucho corroboration that the universe exists independently of ourselves. BUT, we cannot test this notion, so it is not scientific. No observer has been able to observe the universe without him or herself - Zen "no-mind" notwithstanding. But that is a question of philosophy outside the arena of epistemology, and consequently outside the arena of general semantics. We have a nice theory, corroborated by many observers, that the universe is not dependent upon any particular observer, so we abstract and generalize that this extends to the universe less any of these observers. But it is inherently untestable, so it become a simple belief.
Gödel proved that in any sufficently strong system (model, theory, etc.) there are statements that cannot be proven within the system. That applies to belief issues. As long as something remains unprovable, undemonstrable, unfalsifiable, etc., we can choose to assimilate it into a larger system by assuming it to be true, false, or other value. Geometry stands as the paradigm case example. The fith postulate can be choosen in different forms, and each leads to a consistent system - flat (Euclidean), closed curved (spherical), open curved (hyperbolic), etc.
This means that you can chose to join most of us and become a "realist" (of various hue, naive, scientific, sophisticated, etc.), an "anti-realist", an "idealist", etc. and still not conflict with our current scientific "knowledeg" (current accepted theories). It also means you have the same freedom mith respect to religion.
General semantics, in principle, does not answer the question of realism; and it does not answer the question of religion, or any other system based on the hypothesis of an unobserbable.
We probably agree that the universe consists of putative "structures" and that some, but not all these "structures" are observers; that observers have some capacity to "know", "believe", "observer" putative "structures" in the form of abstractions to objects.
The hypothesis proposed holds that
If there are no observers, then there are no structures.
If some structures are observer then there are two possible logical outcomes.
1) there are structures which are not observer.
2) all structures are observers.
I think most of us would agree that 2 is false.
Removing all observers does not remove all structures.
Consequntly the conditional is false.
This is only the metaphysical analysis.
Epistemology cannot answer the question because you took the observer out, and consequently the question moves out of the realm of epistemology.
General semantics, therefore, is silent on the question.
Believe as you choose, enjoy your belief, that is, until it get you into trouble with it not working, then change it.