In both Classical Greek and Latin, there were two and only two sets of personal inflectional endings. Each and every finite verb in Classical Greek and Latin was unconditionally conjugated by using one or the other of these two sets of inflectional personal endings. There were no verbs which were not conjugated in this manner, nor was there ever a third alternative. The Greek and Roman grammarians (Thracians) not only understood this, but they also understood that these two distinct sets of inflectional personal endings represented the very same categories in both languages. For some strange and inexplicable reason those grammarians coming after the Greeks and Romans (non-Thracians) took it upon themselves to take the form of the verb túptomai which was unanimously regarded as the passive form of the verb in Greek and relabeled it the "middle" form of the verb. They did this moreover without changing the labels in Latin. So, in effect, they took a single morphological category "passive" and reinvented it as two separate categories, namely the "middle" and the "passive". In non-Thracian linguistics, changing labels entails changing reality at the same time and thus they were forced to reinvent the Latin language without a middle voice. In non-Thracian linguistics this is simply a matter of making claims such as the following: "In Latin ... the old middle voice 'became' a passive, i.e., was used only in passive functions, while intransitive (/stative) functions were expressed by other morphological means" (Hock 1991: 348). "This active: middle semantic contrast, such as it is, was retained for longer in Greek than in Sanskrit and is totally absent in Latin" (Lightfoot 1979: 241).Then all one needs to do is to repeat such claims three times and they automatically become facts; from there the Woozle effect takes over: "The Woozle effect, also known as evidence by citation, or a woozle, occurs when frequent citation of previous publications that lack evidence misleads individuals, groups, and the public into thinking or believing there is evidence, and nonfacts become urban myths and factoids" (Google).The purpose of this book is threefold: (í) to demonstrate that the set of personal inflectional endings labeled 'passive' in Latin represents the very same category as that expressed by the set of endings labeled 'middle' in Classical Greek, (ii) to present an investigation of all five (sic!) grammatical voices recognized by the Roman grammarians - Palaemon (Keil v: 542.26-27): "And now the principle verbs are actives, pas-sives, neuters, commons, and deponents."Donatus (Keil iv: 383.1-2): "The classes (genera) of verbs, which are called to-kens (significationes) by others, are five: actives, passives, neuters, deponents, and commons."Servius (Keil iv: 413.35-36): "Of verbs there are five classes (genera): actives, passives, neuters, commons, and deponents."Audax (Keil vii: 346.6-7): "How many classes (genera) of verbs are there? Five, that is: active, passive, neuter, common, and deponent."Pompeius (Keil v:227.3): "The classes (genera) of verbs are these: actives, pas-sives, neuters, commons, and deponents."Consentius (Keil v: 367.14-16): "The classes (genera) or tokens (significationes) are five: active as lego scribo, passive as legor scribor, neuter as sto curro, depo-nent as loquor luctor, and common as consolor criminor." -and (iii) to continue the research within Thracian linguistics in order to discover at least two more morphological means of expressing the middle voice in Latin. This will include a form of the verb in Latin corresponding to the so-called 'aorist passive' in Classical Greek which had nothing at all to do with the aorist and is conjugated unconditionally for the active voice.