That there are different forms of observing and that it is important to separate these in one’s mind, is an important theoretical accomplishment.
He who sits in the garden and sees a tree is a first-order observer. He uses his sight and does what he does. Observational systems (not just humans, but also animals, teams, organisations etc.) spend the most time with doing what they do. To be exact, when seeing, one does not ‘know’ during seeing, that one sees. One notices when one is asked about it (“What are you doing there?”) and one then answers (“I am looking at the tree!”). Usually, however, one is then no longer looking at the tree, rather, one thinks and reflects as to why one has not looked at the book one is ‘actually’ reading at the moment for some time.
First-order observations are characterised through their self-understood nature. It is normal to see this (and not something else), to feel, to want, to do. That it could also be different, is, at that moment, not in focus (!), nor is the question why the choice of the focus for the observation was made this way and not another way. To answer these questions or to even ask them in the first place, requires a second-order observer. This can also be oneself.
For change to become possible, it always needs a withdrawal from the self-understood nature of first-order observation. Only when it could also be different, can the focus of the system on the world change.