14 scrupulously pursue the investigation to open a proceeding, not trust that there will always be a “second opportunity” to correct errors, gaps or flaws of the initial investigation, and while this opportunity comes and the State takes advantage of it –if it does occur and the State does in fact take advantage of it- legal security is suspended and justice takes a vacation. 60. We must also review the dies ad quem. We say that measurement of the reasonable term extends until the definitive judgment. Very well, but only in principle. When measuring that term, we must take into account the second instance, when there is one, which may last several months, and sometimes several years. Should we not, then opt for the unappealable judgment, which is the definitive one that cannot be challenged through regular means of recourse? Of course, these measurements must be applied in light of the specific case and taking into account the aspects that European jurisprudence has outlined and that inter-American jurisprudence has adopted, as I mentioned before: complexity of the matter, strategy of the interested party, conduct of the court. XV. Preventive detention 61. Every time the Inter-American Court examines matters such as those of the Case of Tibi, the problem of preventive detention comes up. Certainly, it can arise in connection with reasonable term, which in said conditions should be especially strict and restricted, but also in connection with the very justification of this precautionary deprivation of liberty. Beccaria deemed it to be the punishment before the judgment, and expression that shows the strange nature of preventive detention and its debatable justification. If it is only based on practical reasons (rooted in the inability of justice to find a substitute that at the same time ensures development of the proceeding and security of participants in it, and that enables re-floating of the presumption of innocence), clearly there is a need to contain and contract it: for it to truly be the exception rather than the rule. 62. Despite doctrinal consensus and public discourse on the indispensable reduction of preventive incarceration –which would be another expression of the “minimal” nature of the criminal system in a democratic society, now not only regarding the legal definitions and the punishments, but also regarding the instruments of the proceeding-, what has actually occurred is something different. In our countries preventive detention is liberally applied, in association with systems of prosecution that foster slowness of the proceeding. The number of unconvicted prisoners is very high, as the Latin American Institute for the Prevention of Crime and the Treatment of Offenders (ILANUD), headquartered in San José, Costa Rica, as well as the Inter-American Court, have highlighted. A major part of the effort to further the reform of criminal prosecution –not, incidentally, a “blackboard reform” that functions in the classroom and the seminar, but not in the intractable reality- must have the objective of drastically reducing this army of accused –in other words, “alleged innocents”- who are often a greater number, in the prisons, than their already convicted companions in captivity. 63. Reference to this measure allows us to move ahead in the discussion of other topics that stand out in the set of facts and, of course, in the text of the judgment in the Case of Tibi. Imprisonment was, first, an instrument of retention while the proceeding took place and a judgment was issued. This is the stage reflected in the well-known characterizations of Ulpianus, the Seven-Part Code, and Beccaria himself, already mentioned above: it sought to secure, not punish, the accused, while the trial took place and the judgment was issued. Of course, this careful and compassionate intention was