My presentation shows one insight of philosophical understanding on the learning needs in developing countries, which is often to be abstract and uncommon among countries. The term of learning needs is often stated in policy and projects documents in developing countries. However, the term is often abstract or, if defined, very limited. Moreover, most of developing countries depend on foreign aid, thus they tend to employ the definition of the foreign aid donors, and neglect the country-specific social and cultural needs, such as motivation, teacher-student relationship, and classroom culture. Filling these software needs are crucial as Amartya Sen describes in the capability approach that enable children to have choices to choose, experience, and enact function of opportunities. If the efforts for educational development misconstrue the learning needs, failing to recognized the crucial parts of needs, it may result in creating new problems in equality. In this context, I argue that there are undiscovered learning needs, particularly of software needs, and that these needs are only detected through hearing from adult who obtained information from outside and can compare with others, by recalling their past and conceiving other possible choices and options during their school age.To address this argument, I took a case of Mozambique, and conducted asemi-structured interview on forty adults which I purposefully selected based on age, gender, geographical location of their schools, and profession, through two times’ field trip in February 2016 and February 2017. Mozambique is a low-income country in Southern Africa with having diverse culture, language and ethnic group, and experienced three drastic change of educational system in the history. Thus, taking a case of this country enable me to observe variety of potential learning needs. Based on this data, I present conceptualized learning needs that contains whole picture of learning needs in Mozambique.