Using “Bode plot” to mean a plot of a system’s actual frequency response rather than a straight line asymptotic approximation to the frequency response. (I know – picky, picky. But if it was good enough for Hendrik Bode it’s good enough for me.)
Hendrick Bode may not have had a VNA or SA?
I think it is very likely that he had both, as he worked for Bell Labs.
When I learnt about Bode Plots, back in the early 1980s as part of Telecom Australia's Bridging Training to Tech Officer scheme (The participants proudly called ourselves "Retreads"!
), we were presented with a mathematical expression describing a network, & asked to draw an idealised plot of the amplitude/ frequency & phase/frequency response.
They were, as
ccktek points out, straight line approximations.
People whose "bread & butter" was involved with using swept frequencies to measure amplitude/ frequency response could easily understand the difference between Bode Plots & "real world" results.
On occasion, we used non-swept signal generators to check levels at discrete frequencies across the band of interest, then graphed them manually, if required.
Another "pet peeve" of mine is when people who should know better use the term "in series" to refer to "cascaded" devices!