Nowadays, innovation and translational research concepts are commonly used in science policies. They often refer to the need to contribute to the diffusion of new scientific knowledge toward socio-economic impact. However, novelty is not exclusively related to discoveries: work can remain ignored for a long time, before being recognized. The aim of this work is to examine the mechanisms that could explain the resistance to the circulation of scientific knowledge and its transformation into innovation in the society. It was B. Barber (1961) who was the first in sociology of science to propose an explanation of resistances to discoveries. However, it was S. Cole (1970) who retranslated this question in terms of "delayed recognition" (DR) [1]. In the early 2000s, the problematic of the DR knows a renewed interest in scientometrics with the analysis of "sleeping beauties" (SB), a concept introduced by Van Raan (2004). It refers to an article that goes unnoticed (“sleep”) for more than 10 years ("sleep" period), and then, almost suddenly receives many citations (the "awakening" period) by a ‘‘Prince’’ (PR, another article), attracting a lot of attention from there on in terms of citations (the "Kiss of the Prince") [2]. SBs have been identified in numerous research fields such as biology, chemistry, medicine and physics, and estimated to occur between, 1 and 7.6% [3]. Therefore, study of SB or DR is very interesting in the exploration of mechanisms of circulation of knowledge [4]. The present work aims to explore, through systemic approaches and several case studies, the reasons for DR pattern of citations and what mechanisms are in play in the “awakening” of some scientific discoveries. Our methodology associates quantitative approach, scientometric analysis, and qualitative approaches, historical and sociological analysis, with semi-structured interviews. The first step of our research focused on the scientometric evaluation of the DR phenomena in relation to either (i) the scientific work of Nobel Prize (in Medicine, Physics or Economics), (ii) a concept: “tumor angiogenesis” and "mesenchymal stem cell" (MSC), (iii) a research domain: “astronomy and astrophysics”, (iv) some clinical fields, "oncology" and “drug safety” or (v) applied research: “solid state chemistry”. According to the cases, we extracted from the WoS® database a corpus made up of specific articles (between hundred and a million documents), looked for landmark papers by a historical analysis of article references (RPYS method) [5] and/or filtered DR papers by calculating for all publications the "Beauty coefficient" (B) using the non-parametric criteria of Ke et al. (2015) [3]. For each case studies, we documented the sleeping period and the awakening period by measuring the trends of publications or citations linked to the case, concept and field as well the echo in public media by measuring the press releases in various databases (Google Trends, Factiva). The second step focused on the identification of the Prince, ie the author(s) at the origin of the citation awakening, mainly by co-citations networks analysis. For some cases, we combined this scientometric approach with (i) a historical approach to produce a biography of the author and/or the Prince(s) scientific work to develop a chronology of the scientific discovery using second-hand data, (ii) several semi-structured interviews with researchers who started citing DR papers around (and since), the awakening year in order to grasp the reasons of the delayed recognition. Our scientometric analysis show that DR phenomena could affect article, book or concept, single publication or bundle of publications related to the same discovery, different scientific work related to the same author, single author or group of authors, highly cited scientists or publications as well average H-index author or average cited papers, sometime published in niche scientific journals. The sleeping period could extend form few years to more than half-century, affecting papers published between 1927 and 1988, awakened between 1978 and 2010 and ranging between less than 500 to more than 10.000 total citations. Many reasons have been suggested in the literature as to why the phenomenon of delayed recognition in terms of citations occurs and what mechanisms are in play in the “rediscovery” or the “awakening” of DR papers. Our results show that some publications “sleep” for a long time because they contain scientific claims that steer scientific or ethical controversy in a given field, or because they are at the boundary between two discipline. This “sleeping” period only ends when the controversy is closed, and when consensus is reached. Other publications don’t get enough citations in the beginning because they contain unproven hypotheses, and the citational “awakening” occurs when an experimental proof of the hypothesis is presented. In some applied fields, an industrial application is needed to make a forgotten publication relevant again . In some cases, the DR papers were simply not available to a wide enough scientific community to lack of visibility due to its publication in unknown journals. However, DR cannot be necessarily attributed to the publication of papers in languages other than English. But it could be attributed to the geopolitical situation of some researchers such as Eastern scientists lacking of communication and collaboration with Western scientists that hampered the diffusion of their scientific ideas. In conclusion, the occurrence of delayed recognition raises questions about the relevance of short-term citation-based metrics for the evaluation of scientific impact on which many research public policies are base worldwide. Constancy and continuity in a research field are important components that ensure development of new research subject area. In contrast, lack of continuous financing and research, could hamper the development and growth of new research area. [1] Cole S. (1970), Professional Standing and the Reception of Scientific Discoveries. American Journal of Sociology, 976, 286-306. [2] Van Raan A.F., (2004), Sleeping Beauties in Science. Scientometrics, 59(3), 461-466. [3] Ke Q., et al. (2015). Defining and identifying sleeping beauties in science. PNAS, 112, 7431–7626. [4] Braun T., et al. (2010), On Sleeping Beauties, princes and other tales of citation distributions…. Research Evaluation, 19, 195–202. [5] Bornmann L. et al., (2018), Reference publication year spectroscopy (RPYS) of Eugene Garfield’s publications. Scientometrics, 114, 439–448.