If the universe was essentially unchanging in time, as was generally assumed before the 1920s, there would be no reason that time should not be defined arbitrarily far back. Any so-called beginning of the universe, would be artificial, in the sense that one could extend the history back to earlier times. Thus it might be that the universe was created last year, but with all the memories and physical evidence, to look like it was much older. This raises deep philosophical questions about the meaning of existence. I shall deal with these by adopting what is called, the positivist approach. In this, the idea is that we interpret the input from our senses in terms of a model we make of the world. One can not ask whether the model represents reality, only whether it works. A model is a good model, if first it interprets a wide range of observations, in terms of a simple and elegant model. And second, if the model makes definite predictions that can be tested, and possibly falsified, by observation.

In terms of the positivist approach, one can compare two models of the universe. One in which the universe was created last year, and one in which the universe existed much longer. The model in which the universe existed for longer than a year, can explain things like identical twins, that have a common cause more than a year ago. On the other hand, the model in which the universe was created last year, can not explain such events. So the first model is better. One can not ask whether the universe really existed before a year ago, or just appeared to. In the positivist approach, they are the same.

霍金所说的“宇宙是去年产生的”可不是一个真实的论断，而是一个假定的模型（我们不妨称之为“M1”），结合另一个模型，即“宇宙是更早时间产生的”（命名为“M2”），用以说明“实证法”（positivist approach，或称“证伪法”）的工作原理：

假定我们看到两个长得一模一样的人，我们会问：“他们两个为什么长得这么像呢？”M2可以解释说：“因为这两个人是若干年前由同一个受精卵分裂而成，共享相同的基因，所以长得一模一样。”；而M1，对于不是去年出生的（“that have a common cause more than a year ago”）两个容貌相同的人，就无法给出站得住脚的解释，因为它认定宇宙是去年产生的，那么这两个人一定是随宇宙一同产生的两个分离的个体，那么就不能够证明这两个人“同源”，只能强辩“这是一种巧合”。所以，我们根本不要去管M1和M2哪个更“真实”或更“看起来真实”，单单由于M2能解释M1不能解释的问题，从“实证法”的观点来看，M2就优于M1。

人类认识宇宙的过程正是如此。在人类用肉眼观察宇宙的时候，人们可以“定义”一个宇宙产生的时间（T1），这个时间和人们观察到的宇宙现象完全没有矛盾之处，那么T1在当时就被认为是正确的。而当望远镜发明之后，人们观察到了以前看不见的宇宙现象，有些现象的起源时间（T2）明显早于T1，那么T1就被推翻了。

所以，“实证法”的精髓是，一个模型，不管看起来多么荒谬，只要它能够自圆其说，你就不可以说它是错误的；反之，一个模型，不管看起来多么真实，只要我能举出一个反例，就可以否定它的正确性。“实证法”从根本上不依赖于对模型的“真实性”的考量和讨论，所以从“实证法”的角度来看，“真实”或“看起来真实”都是别无二致的毫无意义。

有两个非常好的例子来说明“实证法”：

1、有种几何学理论，认为平行线是相交的（在无限远处）。这虽然怎么看都不可能是“真”的，怎奈整套理论能够毫不自相矛盾地解释所有的几何原理。

2、爱因斯坦的相对论，看起来匪夷所思，但没人能发现一个反例。

所以，例1和例2，都是完美的模型，至少是在一定范围内的、目前的完美模型。

当然，“实证法”在无“实”可证的情况下也就失效了。比如，人死后是灰飞烟灭还是上天入地，就无法用“实证法”检验。没办法采样。

最后说句题外话：“无罪推定”就是法律范畴内的“实证法”——不管多少人认定你干了坏事，只要找不到证据，你就是无罪的。