Adverbs should be placed as close as possible to the words or phrases that they modify. If you allow an adverb to be separated from the word or phrase that it modifies, the interpretation of the adverb may become ambiguous.
Always place a quantity adverb immediately before the word it modifies.
For all its richness, today's Everglades is a
drastically diminished place.
[Drastically
means to a great extent.]
--Norman Boucher, "Back to the Everglades," Technology Review (modified)
Be especially careful with the placement of the adverbs only and just. Not only can the meaning of a sentence be ambiguous, but the meaning can change drastically as the adverb moves. Consider the following examples.
For the most part, we considered only computer
simulations that mimic the lamprey's neural activity.
[The sentence
suggests that, of various simulations the investigators might have considered,
they focused on just those of a certain sort--perhaps true, but not what the
author wished to communicate.]
For the most part, we considered computer simulations that
only mimic the lamprey's neural activity.
[The
sentence suggests that some computer simulations might do more than mimic--
nonsense, since all simulations just mimic the process they are simulating.]
For the most part, we considered computer simulations that mimic the
lamprey's neural activity only.
[The sentence suggests
that, of more than one feature of the lamprey's behavior or more than one kind
of activity, the investigation singled out neural activity--again possibly true,
but not what the author wished to communicate.]
For the most part, we considered computer simulations that mimic the
lamprey's only neural activity.
[The sentence now
suggests, erroneously, that the lamprey has but a single neural activity and that
this activity was the focus of the investigation.]
For the most part, we considered computer simulations that mimic
only the lamprey's neural activity.
[This is in fact
how the author wrote the sentence, though (at least out of context) the sentence
is still ambiguous. Is the point that the neural activity only of the lamprey,
but not of other animals, was the focus of the investigation, or that only neural
activity was to be simulated?]
--Sten Grillner, "Neural Networks for Vertebrate Locomotion," Scientific American (modified)