Hermann Ebbinghaus introduced the concept of a forgetting curve, i.e., a trace of how the probability of recall changes over time. He used himself as a subject and lists of syllables as learning materials. Other researchers have investigated the same question with real world information. Which of these is the best description of what we learned from these studies:

a. Forgetting curves for both lists of syllables and real information items are negatively accelerated. That is, most of the forgetting happens shortly after the end of learning and the rate of forgetting decreases over time. Eventually, the forgetting curve becomes almost flat, indicating very little further forgetting.
b. Forgetting curves for both lists of syllables and real life information (e.g., Spanish words, class mates from years ago) are positively accelerated. That is, there is very little forgetting in the period shortly after the end of learning but the rate of forgetting increases over time. Eventually, the information is forgotten altogether.
c. Ebbinghaus found negatively accelerated forgetting curves for his syllables lists, but this result only holds for such artificial materials. Studies with real world information (e.g., Spanish words, etc.) found no evidence for such a curve. Instead, forgetting is nearly linear with time; that is, you forget more and more as time goes by, at an approximately constant rate.
d. The forgetting curves for both syllables and realistic information are flat. Once we have taken into account how much was encoded to begin with, it turns out that the probability of recall is the same after a period of time as immediately after learning.