When I read a book I tend to read the words but to also hear them in my mind
Inner Speech is the general name given to this phenomenon,1 although it also goes by imagined speech, silent speech or covert speech, there are few studies and even less definite conclusions as to how it happens and its role in reading comprehension.
and only after that I perceive the whole sentence.
In general, language comprehension goes through the perception of sounds, words and sentences followed by meaning ( independently or in succession) , in order to comprehend a sentence or paragraph you hold words or sentences in some form of short term memory, for visual language ( written words, sign language ), we have a visual sketchpad and for spoken words we have a phonological loop and other inter related memory structures,these buffers also connect or send information to other associational areas in the brain where meaning is presumably derived.2
It seems to me that I can not control this process.
Another related concept to inner speech is private speech3, words spoken out loud to oneself. Private speech is both healthy and normal in children from the ages of 2 to 7 ( a student reading in class to himself in whispers for instance ), although we generally stop the public practice when we grow up due mostly to our social environment and other factors. It is then conceivable that our preferred way ( although not the only one) way of reading for comprehension is to use silent private speech which in turn goes through the language comprehension process outlined above.4
My expectations are that reading has nothing to do with hearing.
Yes and no, while the entry point might be a different one ( visual vs auditory), most likely the language centers and higher associational areas get recruited by either one while reading or hearing, furthermore, by translating internally in between domains, you have access to memories specific to that domain, for instance: when perceiving visually a stop sign (and reacting rapidly), reading and imagining a color, or hearing how something smells5 .
Any idea why this is happening?
Languages consisting of ideograms( Japanese kanji ) present some clues along with the previous overview. Even though the kanji ideograms have more in common with pictures than words; the principal speech comprehension centers get predominantly recruited ( there are still differences though).6 Language has also been predominantly spoken ( the written word is relatively new to our species ) so it makes sense from an adaptive perspective ( i.e it is more efficient) to have a language comprehension center devoted to the older spoken language ( or general sound deciphering) and co-opt it for reading, this new routing happens with the aid of inner silent speech ( which could be a modification of private speech) and other memory buffers.
Notes, reference & sources:
1. A great introduction and overview: Inner Speech: Development, Cognitive Functions, Phenomenology, and Neurobiology
2. Cognitive Neuroscience( Gazzinga) Has a great overview of language and memory and there are whole fields of Neurolinguistics and Psycholinguistics devoted to language.
3. The wiki on private speech and previous sources.
4. From Print to sound and meaning (Cognitive Neuroscience of Language -Kemmerer) presents a model of this arrangement.
5. A contrasting issue is early blind subjects that read in braile, the entry point here is tactile but ends up recruiting language and visual areas the same. see Visual Cortex Activity in Early and Late Blind People
6. Japanese and English sentence reading comprehension and writing systems: An fMRI study of first and second language effects on brain activation Presents a detailed account.