Content uploaded by Elen Le Foll
Author content
All content in this area was uploaded by Elen Le Foll on Jan 24, 2025
Content may be subject to copyright.
Information, Communication & Society
ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/rics20
Giving the outrage a name – how researchers are
challenging employment conditions under the
hashtags #IchBinHanna and #IchBinReyhan
Ahmadou Wagne, Elen Le Foll, Florentine Frantz & Jana Lasser
To cite this article: Ahmadou Wagne, Elen Le Foll, Florentine Frantz & Jana Lasser (20 Jan 2025):
Giving the outrage a name – how researchers are challenging employment conditions under
the hashtags #IchBinHanna and #IchBinReyhan, Information, Communication & Society, DOI:
10.1080/1369118X.2025.2452273
To link to this article: https://doi.org/10.1080/1369118X.2025.2452273
© 2025 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group
View supplementary material
Published online: 20 Jan 2025.
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rics20
Giving the outrage a name – how researchers are challenging
employment conditions under the hashtags #IchBinHanna
and #IchBinReyhan
Ahmadou Wagne
a
, Elen Le Foll
b
, Florentine Frantz
c,d
and Jana Lasser
e,f
a
Department of Informatics, TU Wien, Vienna, Austria;
b
Department of Romance Studies, University of
Cologne, Cologne, Germany;
c
Research platform Responsible Research and Innovation in Academic Practice,
University of Vienna, Vienna, Austria;
d
Technopolis Forschungs- und Beratungsgesellschaft m.b.H., Vienna,
Austria;
e
IDea_Lab, University Graz, Graz, Austria;
f
Complexity Science Hub Vienna, Vienna, Austria
ABSTRACT
Outraged by the release of a ministerial video in which short-term
employment contracts in German academia were lauded through
the embodiment of fictitious doctoral researcher Hanna, thousands
of researchers rallied behind the hashtags #IchBinHanna &
#IchBinReyhan to vent their frustrations about precarious academic
employment in Germany. The emerging connective action attracted
a comparatively large number of researchers in a short period of
time, stayed active for over two years, elicited reactions from
policymakers and the media, and inuenced current legislative
developments in Germany. We analyse the discourse of over 45,000
tweets related to the movement from its onset in June 2021 to
March 2023. Using a mixed-methods approach that combines
machine-learning, corpus-linguistic, and qualitative analysis
methods, we aim to distil the factors that led to the movement’s
considerable success. The fast growth of the movement was likely
driven by the use of an easy-to-personalise action frame and the
large variety of discussion topics, facilitating the involvement of
dierent groups of academics across career levels and employment
conditions. Our analysis of the linguistic characteristics of the
discourse reveals largely positive, constructive, and active
exchanges, in which many of the most salient keywords are lexical
verbs. Our analyses oer an explanation for the continued
involvement of many activists and the successful translation of
solutions developed by the movement to news reports and
proposed law amendments, despite an absence of coordination by
any formal organisation.
ARTICLE HISTORY
Received 7 May 2024
Accepted 8 January 2025
KEYWORDS
Connective action; hashtag
activism; academic working
conditions; IchBinHanna
© 2025 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License
(http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any
medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way. The terms on which
this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent.
CONTACT Jana Lasser jana.lasser@uni-graz.at IDea_Lab, University Graz, Universitätsplatz 3, Graz 8010, Austria;
Complexity Science Hub Vienna, Josefstädterstraße 39, Vienna 1080, Austria
Supplemental data for this article can be accessed online at https://doi.org/10.1080/1369118X.2025.2452273.
INFORMATION, COMMUNICATION & SOCIETY
https://doi.org/10.1080/1369118X.2025.2452273
Introduction
The #IchBinHanna & #IchBinReyhan movement
In 2018, the German Federal Ministry for Education and Research (BMBF) released an
informational video (German Ministry for Education and Research, 2018) on the Federal
Academic Fixed-Term Contract Act (Wissenschaftszeitvertragsgesetz; hereafter Wiss-
ZeitVG), extolling the merits of temporary employment contracts in German academia.
In it, a soft-spoken voice joyfully explains the impact of the law:
This is Hanna. She is a biologist. She is working on her PhD. She has a three-year fixed-
term contract at the university. It has now been extended to three more years. After that,
she cannot sign another fixed-term contract [at a German university] until she has com-
pleted her PhD. […] It’s dierent for Lars, who is a lab technician, and whose employ-
ment contract is therefore governed by regular [German] labour legislation. […] To
ensure that future generations of academics also have the opportunity to gain these qua-
lifications and that not all positions are clogged up by one generation, universities and
research institutions are allowed to grant fixed-term contracts in accordance with the
special rules of the WissZeitVG. This leads to uctuation which, in turn, fosters inno-
vation!
(German Ministry for Education and Research, 2018
1
)
Although the video originally went unnoticed, Hanna later became the infamous symbol
of problematic working conditions in German academia when, in June 2021, the video
was picked up by German early-career researcher, Kirstin Eichhorn, and immediately
went viral on Twitter (now ‘X’). The hashtag #IchBinHanna (#IAmHanna) was initiated
(Bahr et al., 2022) and researchers across Germany took the video as an impetus to vent
their frustrations and anger about precarious academic employment on social media.
Thousands shared their personal experiences of how strings of short, temporary con-
tracts have impacted their professional and personal lives. The hashtag was quickly sup-
plemented by #IchBinReyhan, initiated by Reyhan Şahin, to highlight the intersection of
precarious working conditions with the challenges faced by international and first-gen-
eration researchers. Note that, in the following, we conceptualise #IchBinHanna and
#IchBinReyhan as part of the same social movement and analyse the discourse associated
with these two hashtags together. This is also warranted by the observation that the vast
majority of tweets that contain #IchBinReyhan also contain #IchBinHanna (see sup-
plements for details). The sheer amount of tweets and responses to these hashtags contrived
policy makers to respond, resulting in a debate in the German Parliament just two weeks
after the onset of the movement (Deutscher Bundestag, 2021). It further triggered changes
to the Higher Education Act (Hochschulgesetz) in Berlin (Djahangard, 2021) and, at the fed-
eral level, proposals for changes to the WissZeitVG (Bundesministerium für Bildung und
Forschung, 2023a; Entwurf Eines Gesetzes Zur Änderung Des Befristungsrechts Für Die
Wissenschaft, 2024) evidencing its success in creating political impact.
The #IchbinHanna & #IchBinReyhan movement challenges the individualisation of
systemic problems. The aordances of Twitter debates invited a form of protest that
builds on collectivising these individualised experiences and making them public. The
sheer amount of shared comments made it dicult to ignore or dismiss them, which
was also tangible in the large media coverage that the online discourse has attracted
(see for example Bahr et al., 2022; Haug, 2022; see also Bahr et al., 2023 for a
2 A. WAGNE ET AL.
comprehensive list). At the time of writing, the movement had not reached one of its
main goals, namely making tenured employment after completion of the PhD the rule
rather than the exception. Nevertheless, #IchBinHanna & #IchBinReyhan managed to
actively involve a large number of people for an extended period of time (over two
years), elicited reactions from relevant policy makers, and was crucial in inuencing a
number of legislative texts. As such, it achieved many of the hallmarks commonly associ-
ated with a successful social movement (Bennett & Segerberg, 2012; Tilly, 2019).
With our analysis, we aim to leverage the digital traces of the movement on Twitter to
gain a deeper understanding of the factors that contributed to the movement’s success.
Which factors allowed the movement to scale up quickly and reach as many people as
it did? And how did the movement succeed in remaining active and relevant for a sus-
tained period of time? While there are other accounts of the #IchBinHanna & #IchBin-
Reyhan movement, these contributions either summarise the existing empirical research
on working conditions in German academia (Bahr et al., 2022; Fritz et al., 2021) or con-
tribute to the discussion of possible ways to improve working conditions (Bahr et al.,
2021; Dirnagl, 2022; Fritz et al., 2022). Here, we provide the first empirical analysis of
the content and dynamics of the #IchBinHanna & #IchBinReyhan movement on Twitter
itself.
Working conditions in German academia
Sparked by the informational video on the WissZeitVG featuring ‘Hanna’, the #IchBin-
Hanna & #IchBinReyhan movement focused its criticism on working conditions in Ger-
man academia. Reforms of the German academic system have led to a steady reduction of
permanent positions below the level of full professor: from 2005 to 2018, the number of
academics under the age of 45 working on temporary contracts has increased from 86%
to 92% (Wissenschaftlicher Dienst des Deutschen Bundestages, 2022). This number
includes doctoral researchers, of whom 61% are employed by universities or other
research organisations (Bundesministerium für Bildung und Forschung, 2021) while
the remaining 39% rely on income outside the academic sector to fund their PhD.
Among all researchers employed in German universities, only 9% are professors
(Kreckel, 2016).
Even though precarious working conditions are a pressing concern in many countries,
the situation in Germany is arguably more extreme than in other European countries of
comparable size, as permanent positions other than professorships are extremely rare
(see Kreckel, 2016 for detailed comparisons with France, the UK and the USA). More-
over, in German academia, one in four temporary contracts has a duration of less
than 12 months (Wegner, 2020). The resulting bottleneck situation, caused by a high
number of fixed-term junior positions and too few tenured senior positions, increases
the pressures on individuals (Müller, 2012). As a result, among researchers on temporary
contracts, there are high levels of competition to secure one of the few available tenured
positions (Fochler et al., 2016). An important aspect of the current legal situation in
Germany is that, according to the WissZeitVG, after completion of their PhD degree,
researchers can only be employed on temporary contracts for a maximum of six years
(nine years in medicine). While the initial intention of this law was to force research insti-
tutions to eventually employ researchers on permanent contracts, the reality in German
INFORMATION, COMMUNICATION & SOCIETY 3
academia is that research institutions prefer to employ new early career researchers on
temporary contracts over giving more senior researchers permanent positions. As a
result, researchers who, after six years of fixed-term post-doc contracts, have not yet
landed one of the rare permanent positions have to quit academia or move abroad if
they want to stay in academia. This further increases the pressure on researchers on tem-
porary contracts.
A major rationale for the necessity of temporary contracts rests on the idea that inno-
vation requires a constant inow of new people, bringing new ideas and methods (Wis-
senschaftlicher Dienst des Deutschen Bundestages, 2022). It builds on experiences from
the second half of the twentieth century, where tenured sta were perceived as ‘clogging
the system’ in Germany, with few employment opportunities for emerging scholars
(Kreckel, 2008). However, there is no empirical evidence for the claim that keeping
researchers in precarious working conditions is conducive to the production of more
creative and innovative research (Wissenschaftlicher Dienst des Deutschen Bundestages,
2022). This holds particularly true for high-risk, high-gain and long-term projects that
are less feasible in the current employment conditions (Gibbs, 2015). This line of argu-
mentation, moreover, has been called into question by multiple researchers who have
made clear that precarious working conditions create perverted incentive structures
(Bössel et al., 2023), a fertile ground for exploitation (Lasser et al., 2021), the tacit exclu-
sion of minoritised researchers (Goetze et al., 2023; Lasser et al., 2021; Sørensen & Tra-
week, 2023), alter the very conduct of researchers (Sigl, 2016), and hamper good scientific
practice (Bradler & Roller, 2023).
Connective action
Germany is a country with strong labour rights (Labour Rights Index 2022, 2022),
which makes it all the more surprising that academic workers have been unable to
improve their situation despite years of criticism. Traditional movements organised
along collective action (Olson, 2003) that build on centrally organised protests and
lead figures or organisations struggle to cope with the complex distribution of respon-
sibilities across state and federal actors. This fragmentation has made it challenging to
identify a target of issue advocacy and has undermined solidarity eorts (Morgan &
Pulignano, 2020; Rizzo & Atzeni, 2020). Moreover, organising academic employment
situations along temporary contracts is an expression of the prevalence of neoliberal
and managerial logics that have fundamentally reshaped academia in recent decades
(Sørensen & Traweek, 2023). Researchers have to win at the ‘competitive excellence
game’ before they are granted a long-term perspective (Felt, 2009; Müller & De Rijcke,
2017). Researchers seem to subjectify external demands as individual challenges (Sigl,
2019) and align their practices with the changing requirements (Burrows, 2012).
When individuals fail to secure a permanent position and leave academia, this is
often framed as an individual failure of someone who simply could not keep up
with the high demands of the sector. There has been much less public discussion
of unrealistic ‘excellence’ standards and of the unethical systems operating in acade-
mia that often lead to these decisions (Müller, 2012).
In this article, we argue that what has made the #IchBinHanna & #IchBinReyhan
movement eective in raising concerns about academic working conditions and
4 A. WAGNE ET AL.
steering policy responses is rooted in its characteristics as connective action (Bennett &
Segerberg, 2012; Papacharissi, 2015; Theocharis, 2015). This form of contentious poli-
tics does not rely on hierarchically organised protest, but becomes powerful by giving
voice to a variety of personalised experiences (Bennett & Segerberg, 2012; Olson, 2003).
Collective, decentralised, and individual content sharing allows for complex critiques of
societal issues, relying on a ‘personalised action frame’ (Bennett & Segerberg, 2012) of
activists.
Central to the large-scale and sustained use of an easy-to-personalise action frame are
the aordances of social media platforms, such as hashtags, that enable wide-spread hori-
zontal communication that becomes the central steering element of this form of protest.
The use of hashtags allows social media users to collectively challenge the narratives and
framings of the status quo (Jackson et al., 2020). The narrative agency (Yang, 2016), i.e.,
who can contribute to shaping the story and whose voice is heard in the debate, is not
restricted by institutional aliation or professional status. Despite being condensed
into relatively few characters, hashtags can convey a complex political message. For vul-
nerable or marginalised groups, pushing political messages through the intensified use of
a certain hashtag has presented an opportunity to advocate for social change and voice
powerful counter-narratives, e.g., Occupy Wall Street (DeLuca et al., 2012; Papacharissi,
2016; Theocharis et al., 2015), the Arab Spring movement (Papacharissi, 2016), #MeToo
(Gallagher et al., 2019), #BlackLivesMatter (Yang, 2016), and the hijacking of the
#myNYPD hashtag (Jackson & Foucault Welles, 2015). What has made many of these
protest movements so powerful is their grounding in deeply personal, emotionally raw
messages and experiences that resonate among many people. While academics may
not seem to be a prime example of a marginalised group, being able to access the dis-
course regardless of institutional aliation and professional status is an important aspect
in the context of eeting employment relationships. Interestingly, while precarious work-
ing conditions are usually thought to undermine collective action (Rizzo & Atzeni, 2020),
in the context of #IchBinHanna & #IchBinReyhan, ‘precarity’ itself emerges as the master
frame for collective identity building (Vatansever, 2023).
Whilst the #IchBinHanna & #IchBinReyhan movement has by far been the most
impactful so far, it came after a succession of similar social media movements on Twit-
ter characterised by dierent hashtags, some of which were initiated by the same actors
as #IchBinHanna (Bahr et al., 2020). These movements include #FristIsFrust, which was
initiated in early 2019 and brought to light the frustrations of working on strings of
temporary contracts in German academia (Netzwerk für Gute Arbeit in der Wis-
senschaft, 2019), and #95vsWissZeitVG, an initiative that compiled 95 theses against
the WissZeitVG following Luther’s 95 Theses and was launched in November 2020
(Bahr et al., 2020). The hashtag #ACertainDegreeOfFlexibility began to be used shortly
after #95vsWissZeitVG in early 2021, providing sarcastic takes on the exibility of
employment conditions and place of work that are expected of early career researchers
(Leinfellner et al., 2023). Lastly, #DauerstellenFürDaueraufgaben, which was launched
in early 2022 with strong support from relevant German trade unions. It demanded
permanent contracts for academic sta involved in permanent activities such as teach-
ing. As evidenced by these predecessors, digitally networked action is not guaranteed to
scale up and become a stable movement that is capable of targeted action and inuen-
cing policy changes. The success of an action involves ‘complex rhetorical action of
INFORMATION, COMMUNICATION & SOCIETY 5
activists as “bits and pieces” that eventually become a collective identity that strangers
(as well as friends and allies) want to embody through their actions’ (Alfonzo & Foust,
2019).
As such, the #IchBinHanna & #IchBinReyhan movement oers an interesting
example of a connective action that allows us to study the aspects that have led to its con-
siderable success. In this context, we aim to answer the following research questions: (1)
What dierentiates #IchBinHanna & #IchBinReyhan from preceding Twitter campaigns
such as #FristIstFrust and #95vsWissZeitVG? (2) To what extent did diverse groups of
social media users contribute to this connective action? (3) What characterises the
language used by dierent groups of social media users in tweets belonging to the #Ich-
BinHanna & #IchBinReyhan discourse?
Materials and methods
To answer these research questions, we employed mixed methods analysis of tweets
posted under the hashtags #IchBinHanna and #IchBinReyhan. Our various approaches
aim to complement each other, enabling us to triangulate our findings from more
than one perspective. In the following, we describe the data collection and data proces-
sing, computational and qualitative analysis.
We used public Twitter data, exempted from ethics review under our research insti-
tutions’ rules. In the ‘Keyword Analysis’ section, we mention the handles of key accounts
involved in the #IchBinHanna & #IchBinReyhan discourse. Most accounts are insti-
tutions or public figures such as politicians and journalists, but some are private individ-
uals. We only include private accounts that were actively contributing to the discourse.
We argue that it is ethically admissible to list these accounts (Townsend & Wallace,
2016), as mentioning their handles in this work will neither significantly heighten the
attention they receive nor breach their expectations.
Data collection
We compiled a corpus of tweets related to the #IchBinHanna & #IchBinReyhan move-
ment on Twitter (now ‘X’), using the academic search endpoint of the Twitter API
and twarc2 (Summers et al., 2022). We collected all tweets that contained the hashtags
#IchBinHanna, #IchBinHannah or #IchBinReyhan (in all forms of capitalisation)
between 1 June 2021 and 30 March 2023. Note that we analyse #IchBinHanna and #Ich-
BinReyhan as part of the same movement, highlighting dierent but related aspects of
precarious working conditions in academia. This was motivated by the observation
that the two hashtags frequently co-occur in tweets and that emerged around the same
time. However, we acknowledge that the two hashtags emphasise dierent lived experi-
ences in academia. We combined the tweets collected for both hashtags in a single
corpus.
To compare the #IchBinHanna & #IchBinReyhan movement to similar Twitter-based
movements concerned with the working conditions of academics in Germany, we also
collected tweets for the hashtags #95vsWissZeitVG (Bahr et al., 2020), #FristIstFrust
(Netzwerk für Gute Arbeit in der Wissenschaft, 2019), #ACertainDegreeOfFlexibility
6 A. WAGNE ET AL.
(Leinfellner et al., 2023), and #DauerstellenFürDaueraufgaben (and common variants)
between 1 January 2019 and 4 February 2023.
Data was collected at irregular intervals. We collected all new tweets up to the time of
data collection on 26 June 2021, 10 November 2021, 14 April 2022, and 4 February 2023
for all hashtags, and on 30 March 2023 for #IchBinHanna and #IchBinReyhan only. Data
collection was halted after 30 March 2023 when the Twitter data access API became
defunct. We excluded retweets and retained only tweets in German and English, as ident-
ified by tweet metadata. For each of the hashtags, we report the number of unique users
and overall number of tweets using a hashtag, as well as the average number of tweets per
day in Table 1.
We focus on Twitter discourse. We do not collect data from any other platforms like
Facebook or Instagram. Despite this limitation, most of the relevant social media dis-
course occurred on Twitter, as highlighted by numerous media references (Bahr et al.,
2023). In addition, analysis using Facebook’s CrowdTangle revealed only 399 posts
(and 1,298 comments) with the hashtags #IchBinHanna & #IchBinReyhan between 1
June 2021 and 30 March 2023. Even though CrowdTangle is limited since it only covers
pages with over 25,000 followers, and groups with more than 95,000 members, these
figures pale compared to the 45,715 tweets mentioning the hashtags, demonstrating
that discussions were largely concentrated on Twitter.
Data processing
We first removed line breaks, punctuation marks, URLs, and tokens that only consisted
of numbers using the BeautifulSoup library (Richardson, 2007) and regular expressions
in Python. The tweets were then converted to lowercase and tokenised using the Tweet-
Tokenizer from the NLTK library (Bird et al., 2009). Next, tokens were lemmatised to
retain only the uninected headwords (e.g., prekäre Arbeitsbedingungen became prekär
arbeitsbedingung) using the WordNetLemmatizer of the NLTK library (Bird et al.,
2009) for the English tweets, and the HannoverTagger (Wartena, 2019/2023) for the Ger-
man ones. Stopword lists, consisting mostly of function words such as articles, pronouns,
and auxiliary verbs, were derived from the most frequent lemmas (grouping of all
inected forms of a word) as extracted from two large web corpora: the DeTenTen20
for German and the EnTenTen20 for English (Jakubíček et al., 2013). These stopwords
were then removed, together with lemmas consisting of two or fewer characters. All sub-
sequent quantitative analyses were carried out on this pre-processed corpus consisting
only of the content lemmas.
Table 1. Descriptive statistics of actions.
Hashtag Number of unique users Number of tweets Average nb. of tweets per day
IchBinHanna & IchBinReyhan 8,289 45,715 69.5
95vsWissZeitVG 1,404 6,195 15.7
FristIstFrust 1,263 5,379 5.3
ACertainDegreeOfFlexibiity 1,092 3,818 11.9
DauerstellenFürDaueraufgaben 106 425 1.9
Note: Number of unique users, unique tweets, and average number of tweets per day for the hashtags #IchBinHanna,
#95vsWissZeitVG, #FristIstFrust, #ACertainDegreeOfFlexibility, and #DauerstellenFürDaueraufgaben.
INFORMATION, COMMUNICATION & SOCIETY 7
Account classification
To analyse the content posted by specific user groups, we classified the 8,289 unique
Twitter accounts featured in our corpus. First, we removed user accounts that only con-
tributed tweets not identified as either ‘German’ or ‘English’ by tweet metadata, leaving
7,877 unique accounts with 7,943 unique usernames. The dierence between the number
of accounts and usernames occurs because users can change their account name. Treat-
ing the same account but with dierent usernames as distinct allowed us to assign
accounts to dierent user groups at dierent points in time, if their account name
(and profile description) was consistent with another group.
Based on an initial qualitative analysis of the accounts’ profile descriptions, we ident-
ified six groups of academics that participated in the discussion: students, doctoral
researchers, postdoctoral researchers, junior professors (equivalent to assistant pro-
fessor), full professors, and academics who do not disclose their status (academic unspe-
cified). In addition, we identified several groups of users that are not personal accounts or
have non-academic backgrounds: institutions, media, union representatives, political
representatives, teachers, medical doctors, and bots. We subsequently devised a heuristic
that automatically assigns accounts to one of these 13 groups based on keywords such as
‘student’ or ‘professor’ featured in the account names and profile descriptions. Accounts
that could not be associated with any group by the automated classification heuristic were
labelled as unclassified. To validate this approach, we drew a random sample of 2% of all
unique user accounts featured in our corpus (162 accounts) and tasked two human anno-
tators (both authors of this publication) to manually assign labels to each account in the
sample. The interrater agreement between the two human annotators and the automated
annotation (Fleiss’ Kappa 0.67) was not as high as the agreement between the human
annotators (Fleiss’ Kappa 0.84), nonetheless high enough to warrant the use of the auto-
mated classification heuristic to automatically assign group labels to user accounts. With
this classification scheme, just over half of user accounts (50.8%) remain unclassified after
automated classification. This is due to the fact that many accounts provide very little
information in their profile. The number of accounts that remain unclassified in the
manually annotated samples is equally high. However, unclassified accounts contribute
only 28.6% of the total tweet volume.
Topic modelling
We use the tweet corpus together with a list of known events and media coverage (Bahr
et al., 2023) to identify and describe a comprehensive list of events relating to the #Ich-
BinHanna & #IchBinReyhan movement. We provide details of our computational
approach for event detection in the Supplement. Of the identified events, we chose
four that were of particular interest: (1) the initiation of the movement on 10 June
2021, (2) the debate on the topic in the German Bundestag on 24 June 2021 (Deutscher
Bundestag, 2021), (3) a Zoom conference of the German trade union GEW on 1 July 2021
(Gewerkschaft Erziehung und Wissenschaft, 2021), and (4) the publication of a draft
amendment of the WissZeitVG on 17 March 2023 (Bundesministerium für Bildung
und Forschung, 2023b). The choice of events is motivated by the high volume of tweets
associated with them, the diversity of their contexts (discussion in the German Bundestag
8 A. WAGNE ET AL.
vs. discussion hosted by a union), and their distribution over time (beginning of the
movement in 2021 vs. the announcement of the draft amendment in 2023).
To characterise these events in detail, we aimed to detect latent topics in the tweet cor-
pus that were posted in the time periods corresponding to the events. To this end, we
performed topic modelling using BERTopic (Grootendorst, 2022). BERTopic first trans-
forms each document into a vector of numeric values–a process also called ‘embedding’–
and then looks for clusters of documents (i.e., ‘topics’) based on their distance in the
embedding space. BERTopic relies on pretrained BERT-based transformer models (Dev-
lin et al., 2018) to create the embeddings. Since our corpus contains both German and
English tweets, we used a multilingual sentence transformer model to embed documents
(Reimers & Gurevych, 2019). Due to the curse of dimensionality, we used UMAP
(McInnes & Healy, 2021), a non-linear dimensionality reduction approach to reduce
the embeddings to two dimensions, prior to clustering. We then used HDBSCAN
(McInnes et al., 2017), a density-based hierarchical clustering approach to identify clus-
ters of documents. For both UMAP and HDBSCAN, we used the default parameters pro-
vided in the Python implementation of BERtopic (Grootendorst, 2022) except for the
language setting, which was set to ‘multilingual’. HDBSCAN optimises for the coherence
of clusters while keeping the number of clusters minimal. In contrast to other clustering
approaches like K-means, HDBSCAN works without specifying the number of expected
clusters in advance and automatically identifies the optimal number of clusters. The out-
put of HDBSCAN is a mapping between documents and clusters, where each document
is assigned to exactly one cluster, which we interpret as the ‘topic’ of the document.
We fitted one topic model for each of the four events. For each topic, we then extracted
words that are representative of the topic based on all documents assigned to the topic, using
a class-based term frequency-inverse document frequency approach. The output of the topic
model is therefore a list of n topics that is described by 10 representative words each, where
the number of topics is determined automatically. This allowed us to qualitatively interpret
the events by examining the representative words for each topic and inspecting individual
tweets assigned to the topic (see also Section ‘Qualitative analysis’ below).
Keyword extraction
To triangulate the results of the topic models and provide an analysis of the linguistic
characteristics of the whole discourse, we employed a method commonly used in cor-
pus-assisted discourse analysis: keyword analysis (e.g., Gabrielatos, 2018). It involves
comparing the frequency of words in a target corpus to their frequency in a reference
corpus using various statistical measures of association to measure their ‘keyness’.
Terms with a high keyness score are more likely to be associated with the target corpus
and can therefore be considered keywords. Unlike topic modelling, this method cannot
be applied to multilingual corpora; hence, our keyword analysis focused exclusively on
the German subset (36,353 tweets, i.e., 87.5%) of our full corpus.
To perform keyword analysis, we compiled a reference corpus consisting of a random
sample of over three million unique German language tweets. This was achieved by
extracting all German-language tweets that were included in the 1% Twitter sampled
stream posted within the same time frame as the target corpus (Lasser, 2022; Pfeer
et al., 2023). These tweets were pre-processed following the same procedure as for the
INFORMATION, COMMUNICATION & SOCIETY 9
German #IchBinHanna subcorpus (see Section ‘Data processing’ above). We then
applied the textstat_keyness function of the quanteda R package (Benoit et al., 2018) to
identify potential key lemmas. We chose the likelihood ratio G
2
statistic as the measure
of association between the target corpus and the reference corpus and applied William’s
correction to reduce the potential for Type 1 error (Benoit et al., 2018). The function
returned a list of lemmas ordered by their keyness score (G
2
), of which 6,521 were sig-
nificantly associated with the target corpus at the significance level of 0.001. However,
given that G
2
values are very sensitive to corpus size, we opted for the Bayesian Infor-
mation Criterion (BIC) for keyword selection, which takes account of the total number
of words in the compared corpora (Wilson, 2013). BIC scores >10 are said to ‘provide
very strong evidence against H
0
’ (Gabrielatos, 2018) and this was therefore chosen as a
minimum threshold. In addition, an arbitrary minimum of 50 occurrences across the
entire German #IchBinHanna subcorpus was set to exclude idiosyncratic keywords
used only by a minority of Twitter users. These steps narrowed down the list to a man-
ageable subset of 613 keywords for the subsequent qualitative analysis.
Qualitative analysis
Following a computational grounded theory approach (Nelson, 2020), we relied on qualitat-
ive analysis to validate patterns identified via computational analysis (i.e., event detection,
user classification, topic modelling, and keyword analysis) and interpret our quantitative
findings. Our personal understanding of the #IchBinHanna & #IchBinReyhan discourse
also played a significant role in our analysis, as two authors were involved in the movement,
followed the discourse over an extended period, and participated in associated events.
As described above, we first used inductive computational text exploration methods
with unsupervised machine learning (topic modelling) and word scores (keyword extrac-
tion) for pattern confirmation (Nelson, 2020). For topic modelling, we experimented
with parameters to avoid overly broad or granular topics, grounding these decisions in
our understanding of the discourse. Finalised parameters for topic modelling were the
default options for BERTopic. This knowledge also guided our cut-o threshold selection
process for examining relevant keywords.
Because of the large corpus size, we used computation-guided reading to avoid poten-
tial bias from researcher-led document selection. Events were identified through tweet
counts over time, supplemented with known event lists associated with the movement.
Topic models and user classification were used to select document subsets for further
deep reading. Additionally, we created an interactive table of all tweets and relevant
metadata, searchable by usernames and keywords and including links to their original
source, as a convenient way to compare, examine, and interpret the use of specific key-
words by dierent users and groups of users in context. The second step, pattern
interpretation, involved qualitative deep reading of the documents for a better under-
standing of the arguments, which are often missed by computational methods.
Results
In the following, we present the results of our mixed-methods analysis of the #IchBin-
Hanna & #IchBinReyhan discourse. We follow a computational grounded theory
10 A. WAGNE ET AL.
approach (Nelson, 2020), alternating between unsupervised and corpus-linguistic
methods to analyse the digital traces left by the participants of the discourse and a quali-
tative analysis of events, discussions, and keywords. We start by comparing the #IchBin-
Hanna & #IchBinReyhan movement to other preceding movements. Next, we
characterise the involvement of dierent participant groups in dierent discussion topics
for dierent events. Lastly, we analyse the language used by the activists in their tweets,
looking for linguistic characteristics that could have contributed to the success of the
movement. Following Bennett and Segerberg (2012) and Tilly (2019), we conceptualise
a successful movement as one that manages to engage a large number of people over
an extended amount of time, gets covered by mainstream media, and triggers legislative
changes.
Comparison with other movements
To answer the question of how the #IchBinHanna & #IchBinReyhan movement com-
pares to preceding movements, we collected tweets associated with these movements
for the time they were active on Twitter. Figure 1(panel A) shows the total number of
tweets featuring hashtags related to various movements that broach the issue of precar-
ious working conditions in German academia. The number of unique Twitter users using
these hashtags is shown in Figure 1(panel B). Compared to the #95vsWissZeitVG cam-
paign, #IchBinHanna & #IchBinReyhan mobilised about six times as many users who,
collectively, posted about seven times as many tweets over a shorter period of time.
This is also visible in the number of tweets per day: the #IchBinHanna & #IchBinReyhan
movement produced on average 69.5 tweets per day, whereas #95vsWissZeitVG and
#FristIstFrust generated 15.7 and 11.9 tweets per day, respectively (see also Table 1).
The time series of tweeting activity also shows that the use of other hashtags declined sig-
nificantly after the onset of #IchBinHanna & #IchBinReyhan, suggesting that users con-
verged to the most popular hashtag at the expense of others, making use of the hashtag as
an aordance for participatory coordination (Alfonzo, 2021).
Figure 1. Twitter users’ use of hashtags related to working conditions in German academia between 1
January 2018 and 4 February 2023. A: Time development of tweet volume.
Note that tweet volume is visualised on a logarithmic scale and that a 30-day moving average was applied. B: Number of
unique users for each movement of interest.
INFORMATION, COMMUNICATION & SOCIETY 11
Characteristics of the participants
We also aimed to shed light on the distinct groups of people that have engaged in dier-
ent aspects of the social media discourse. In particular, we were interested in analysing to
what extent groups not directly aected by precarious working conditions (yet), such as
tenured professors and students, showed intergroup solidarity and participated in the
discourse. To this end, we classified the 7,943 unique Twitter accounts that quoted the
#IchBinHanna or #IchBinReyhan hashtags into dierent groups of academics and
non-academic users (see Methods). Figure 2 shows the extent to which dierent groups
contributed to the overall tweet volume. Out of all the accounts that could be successfully
classified, the five most prevalent groups comprise individuals in academia: ‘unspecified
academics’ (1,403 unique accounts, 17.7% of tweet volume from classified accounts),
postdoctoral researchers (901 accounts, 11.3%), doctoral researchers (497 accounts,
6.3%), full professors (458 accounts, 5.8%), and students (224 accounts, 2.8%). As
most junior professors (25 tweets, 0.3%) in Germany have a similar status to postdoctoral
researchers – both in terms of academic hierarchy, as well as in terms of the precarity of
their working conditions – we combined the two groups in a single ‘postdoc’ group going
forward.
We focus the remainder of our analysis on the five most prevalent groups: postdocs,
doctoral researchers, unspecified academics, professors, and students. We report the
results for the remaining groups (institutions, media representatives, union representa-
tives) together with the unclassified accounts as ‘other/unclassified’. We excluded five
accounts that were classified as bots.
Figure 3 shows the number of tweets that were contributed by the five most prevalent
groups over time. Significant trends, including the high volume of tweets during the
movement’s initial phase in June 2021, as well as spikes in activity observed in the sum-
mer of 2022 and March 2023, manifest consistently across various groups. However,
Figure 2. Distribution of user groups in our corpus. The area of each rectangle corresponds to the
volume of tweets published by each group, whilst n corresponds to the total number of unique
Twitter accounts in each group.
12 A. WAGNE ET AL.
some dierences between user group activities are noteworthy: Of particular interest is
the development of the number of tweets published by accounts identified as belonging
to full professors. While this group contributed a substantial number of tweets initially –
more than students and doctoral researchers, their activity continuously decreased until
late 2022, when the group contributed the least number of tweets. However, whilst the
activity of all groups increased sharply in March 2023, the increase in tweet volume
from professors was particularly notable, reecting the emergence of the #ProfsFuer-
Hanna (Professors for Hanna) movement around that time (ProfsfuerHanna, 2023).
Participation of dierent actor groups
To analyse specific groups’ contributions to the discussion, we examine four major events
characterised by an increased volume of tweets: (1) movement onset, (2) German Bun-
destag discussion, (3) GEW conference, and (4) release of the first draft amendment.
Using topic modelling, we identify and describe topics discussed in each event, assigning
tweets to representative topics. Topics are described by their most representative words.
For example, the words Innovation [innovation], fördern [foster], Fluktuation [uctu-
ation] and innovativ [innovative] describe a topic that was discussed during movement
Figure 3. Number of tweets over time posted by the five largest user groups: Full professors (orange),
postdoctoral researchers (yellow), doctoral researchers (dark blue), students (light purple), academic
unspecified (light blue). All remaining user groups are aggregated under the label ‘other/unspecified’
(dark purple).
Note that tweet volume is visualised on a logarithmic scale.
INFORMATION, COMMUNICATION & SOCIETY 13
onset, echoing the content of the original ‘Hanna’ video. We describe events (1) and (4)
in detail here and provide descriptions of events (2) and (3) in the supplement. For each
event, we discuss the most frequent five topics, interpreted in their specific contexts
through deep reading. Table 2 shows the tweet count for these topics and the contribut-
ing user group’s percentage. We provide a full list of topics per event in the Supplement.
Our corpus consists of both German and English language Tweets, with English ones
being less common. English-language topics related to German academia ranked highly
in the initial three events, with full professors significantly involved. During the move-
ment’s onset, their contribution to this topic was five times greater compared to other
topics. Most of these tweets tried to summarise the movement’s discussions for an inter-
national viewership. In addition, each event featured dierent topics concerning aca-
demic work, shaped by the event’s specific context. We further detail these for two
events subsequently.
Onset of the movement
During movement onset, the most prevalent topic is characterised by the words ‘Uni |
studieren | Wissenschaft’, relating to a broad discussion of academia in connection to
higher education and studying, including many users sharing how precarious working
conditions have shaped their experiences as university students. Accounts classified as
students actively contribute to this topic, accounting for 4% of tweets, twice their average
contribution to the discourse. Conversely, postdocs, typically contributing 29% of tweet
volume, are only responsible for 11% of tweets in this topic. The high volume of unclas-
sified accounts (74%) suggests the topic reached a broader audience beyond academic
circles.
After the English language topic, the topic characterised by ‘Wissenschaft | BMBF |
Bund’ emerged as the third most prevalent one. In this topic, the laws governing precar-
ious working conditions in German academia are discussed and the BMBF is identified as
the main responsible actor that could change the status quo.
The fourth most prevalent topic, characterised by ‘gender | studie | studieren’, primar-
ily consists of critical tweets regarding the #IchBinHanna & #IchBinReyhan movement.
Critics claim that the work of academics quoting these hashtags, particularly from huma-
nities and social sciences, is ‘unimportant’ or ‘useless’. (e.g., Physiker sind wichtig für
dieses Land – Gender Studies hingegen nicht. Und unter #IchBinHanna sind eher Letztere
zu finden. [Physicists are important for this country – but gender studies are not. And
the latter are more likely to be found under #IchBinHanna.]). Many argue that research-
ers in these disciplines, notably gender studies, do not deserve permanent contracts or a
living wage. A significant portion (76%) of the tweets are from unclassified accounts,
suggesting that the discourse extends beyond academic circles, with several seemingly
anonymous or ‘troll’ accounts involved.
The fifth most prevalent topic ‘Innovation | fördern | Fluktuation’ critiques the
hypothesis that high workforce uctuation fosters innovation – a key justification
brought forward by actors defending the status quo. Other topics (not shown in
Table 2) cover aspects of precarious working conditions, including being a first-gener-
ation academic, working while unemployed (Erwerbsarbeitslosigkeit), dependence on
project funding, and motherhood. Furthermore, temporary contracts and solutions
like unionisation are addressed.
14 A. WAGNE ET AL.
Table 2. Number of tweets for the five most prevalent topics identified for four events.
First three words characterising a
topic
Number of
tweets
% contributed by
students
% contributed by
doctoral researchers
% contributed by
postdoctoral
researchers
% contributed by
full professors
% contributed by
academics
unspecified
% contributed by
other/unclassified
Onset of the movement on June 10
Uni | studieren | Wissenschaft 299 4 4 11 2 5 74
German | Germany | academic 181 2 10 21 19 20 28
BMBF Bund | Wissenschaft |
WissZeitVG
163 2 2 21 6 16 53
Gender | Studie | studieren 119 1 3 10 2 4 80
Innovation | fördern |
Fluktuation
96 0 6 25 6 23 40
Aktuelle Stunde
Debatte | Diskussion |
Argument
75 3 4 40 4 17 32
German | Germany | academic 71 1 6 25 8 27 33
Vertrag | Promotion | befristet 60 5 3 32 12 8 40
Danken | AnjaKarliczek |
HannaImBundestag
50 0 2 32 12 16 38
Wissenschaft |
WissenschaftlerInnen |
AnjaKarliczek
49 2 18 31 2 10 37
GEW conference
Doktorvater | Gespräch |
DoktorandInnen
68 1 7 24 18 19 31
Thread | sehr | lesenswert 52 4 8 27 6 31 24
German | academia | Germany 37 0 8 54 11 19 8
Ministerin | Anjakarliczek |
realistisch
35 0 3 11 14 23 49
Kolumne | JanBoehm |
IchBinHanna
32 0 6 34 3 28 29
Publication of draft amendment
Professur | uni | Jahr 136 1 5 33 10 18 33
(Continued )
INFORMATION, COMMUNICATION & SOCIETY 15
Table 2. Continued.
First three words characterising a
topic
Number of
tweets
% contributed by
students
% contributed by
doctoral researchers
% contributed by
postdoctoral
researchers
% contributed by
full professors
% contributed by
academics
unspecified
% contributed by
other/unclassified
Researchwonderland |
wisssystemfehler | WissZeitVG
76 1 4 49 3 17 26
Wissenschaft |
Wissenschaftlerinnen | BMBF
Bund
70 1 7 16 4 23 49
WissZeitVG | bambule |
ChatGPT
69 3 3 45 9 7 33
Jahr | fordern | Postdocphase 52 2 2 27 0 29 40
Note: For each topic we report the percentage of tweets contributed by users identified as students, doctoral researchers, postdocs, and professors, as well as accounts identified as ‘academics’
without further specification, and accounts belonging to other groups as well as unclassified accounts. We exclude tweets by accounts identified as ‘bot’.
16 A. WAGNE ET AL.
In general, topics discussed during movement onset are diverse, describing dierent
aspects of precarious working conditions and how they intersect with other forms of dis-
crimination such as ableism, sexism, racism and classism. While most topics outline pro-
blems, some identify actors like BMBF and unions for potential solutions. The initial
Twitter discourse shows diverse participation: Postdocs – who contributed 29% of all
tweets over the course of the movement – only contributed 14% of the tweets in this
initial phase, while full professors contributed 7%, and unclassified or other groups con-
tributed 63%.
Release of a draft law amendment
The second event we describe here is the discourse around the proposed draft amend-
ment of the WissZeitVG on 17 March 2023. The proposal aimed to reduce the maximum
duration of temporary contracts after completion of the PhD degree from six to three
years (Bundesministerium für Bildung und Forschung, 2023b). The primary focus of
the discussion, highlighted in the most prevalent topic ‘Professur | Uni | Jahr,’ revolves
around the potential impact of this change on completing a Habilitation, a requirement
for full professorship for many research disciplines in Germany.
The second prevalent topic mentions two new hashtags, #WissSystemFehler and
#ResearchWonderland, emerging in 2022 and 2023, respectively. #WissSystemFehler is
associated with specific dysfunctions in the academic system, like data fabrication and
research funding embezzlement. #ResearchWonderland sarcastically refers to a BMBF
public relations campaign aimed at attracting international researchers to Germany
(Goetze et al., 2023). Notably, postdocs contribute significantly, making up 49% of the
tweets in this topic.
The third and fourth most prevalent topics (‘wissenschaft | BMBF Bund | Tag’) and
(‘WissZeitVG | Bambule | ChatGPT’) overlap and both refer to the organisation of a pro-
test in front of the ministry in reaction to the publication of the draft amendment. The
second topic additionally ridicules the quality of the draft amendment, with a substantial
number of tweets suspecting that it had been written using ChatGPT.
The fifth topic discusses the bill’s details and emphasises key points regarding the
establishment of permanent contracts for postdocs (‘Jahr | fordern | postdocphase’).
Remarkably, there are no tweets from accounts identified as professors in this topic. In
contrast, the sixth most prevalent topic, ‘Professorinnen | profsfürhanna | Profs’ (not
shown in Table 2), marks the emergence of the #ProfsFürHanna movement. This move-
ment, uniting professors in support of #IchBinHanna gained substantial traction in
March 2023, with a 10% contribution from professorial accounts.
Linguistic characteristics of the discourse
Lastly, we turn to the results of a corpus-linguistic analysis of the language of the tweets
that contributed to the discourse. Using keyword analysis, we extracted 613 lemmas that
are strongly associated with the German tweets in the corpus (see Methods for details and
code repository for full list (Wagne et al., 2024)). Among these lemmas, we find over 70
hashtags (see Figure 4(panel A) for a selection). Hashtags associated with comparable
movements (e.g., #FristIstFrust, #95vsWissZeitVG) are both frequent and specific to
the #IchBinHanna discourse, indicating their topical closeness. On the other hand,
INFORMATION, COMMUNICATION & SOCIETY 17
hashtags such as #karrieren, #Bundestag, #Bildung and #BTW21 (the latter referring to
the German federal elections in 2021) are less specific to the discourse, yet still occur with
a high frequency.
Keyword analysis also allows us to identify key actors of the movement. Figure 4(B)
shows a selection of the mentioned Twitter handles most strongly associated with the dis-
course (the full list features 64 handles). Unsurprisingly, the three initiators of the move-
ment, @amreibahr, @drkeichhorn and @sebastiankubon, feature prominently. The
analysis also unveils key actors that do not emerge from the analysis of tweet frequency
by group, such as @anjakarliczek – then Federal Minister for Education and Research,
@gew_bund – the account of the GEW, and @bmbf_bund – the account of the BMBF.
These accounts are frequently mentioned but more rarely actively contribute themselves.
Interestingly, and in contrast to other digitally networked movements like Black Lives
Matter (Alfonzo, 2021), the majority of key actors are politicians, institutions, unions
or media accounts and journalists. Even though #IchBinHanna & #IchBinReyhan
never became an organisationally enabled network (Bennett & Segerberg, 2012), this
points towards considerable eort in directing communication towards institutional
actors that activists identified as having power to change the situation.
Notably, many of the most strongly associated keywords are lexical verbs, many of
which point to agency in the language used by the activists (Formanowicz et al.,
Figure 4. Visualisation of selected keywords associated with the German-speaking #IchBinHanna &
IchBinReyhan Twitter discourse: A hashtags, B mentioned handles, C verbs, and D adjectives and
adverbs.
Note that the hashtags used as selection criteria for the data collection (see Methods) were excluded from panel A.
18 A. WAGNE ET AL.
2017). These include verbs describing the status quo (e.g., leisten, [Drittmittel] einwerben,
bewerben, verlassen [achieve, secure [third-party funding], apply, leave/quit]), as well as
verbs describing alternatives and ways forward (e.g., diskutieren, reformieren, verbessern,
finanzieren [discuss, reform, improve, fund]). A selection of the verbs most strongly
associated with the discourse is displayed in Figure 4(C). The four most characteristic
and most frequent verbs are promovieren [to do a PhD], entfristen [to make a temporary
contract permanent], forschen [to do research] and verstopfen [to clog]. The latter was
featured in the initial Hanna video in the context of ‘permanent positions clogging up
the system’ (see Introduction). This wording outraged many academics and was there-
fore widely taken up in response tweets. Aside from entfristen, the full list of keywords
features numerous other words derived from the word Frist [period of time, deadline],
including befristet [temporary], Befristung [fixed-term employment], unbefristet [perma-
nent], #FristIstFrust [#TemporaryIsFrustrating], Entfristung [to make a temporary con-
tract permanent], Dauerbefristung [continuous temporary contracts], #EntfristetHanna
[#MakeHannaPermanent], langfristig [long-term], Befristungspraxis [the practice of tem-
porary contracts], befristen [to make employment temporary] and Kettenbefristung
[string of temporary contracts], reecting the fact that discussions often revolved around
temporary contracts and their impact on researchers.
Lastly, Figure 4(D) shows a selection of the adjectives and adverbs most character-
istic of the discourse. We qualitatively analysed the sentiment of these keywords: the
colours indicate whether these lemmas are mostly associated with positive (green),
neutral (grey) or negative (red) sentiments. Unsurprisingly, befristet [temporary]
and unbefristet [permanent] are both very frequent and characteristic, and are associ-
ated with opposing sentiments. Similarly, the word prekär [precarious] carries nega-
tive sentiment and is also both highly characteristic and frequent – a trace of the
collective identity building of activists around their status as academic precariat
(Vatansever, 2023). Nonetheless, the graph shows that contributors to the discourse
did not exclusively vent frustrations and/or criticise the status quo, but also used posi-
tively connotated adjectives and adverbs to suggest alternatives and encourage each
other (e.g., wichtig, lesenswert, gemeinsam, möglich [important, worth reading,
together, possible]). The prominence of keywords such as strukturell, oft, ständig
and häufig [structural, often, always, frequently] hints at the fact that the experiences
shared in the discourse, even though presented anecdotally in the first person, rep-
resent systemic issues in German academia.
Discussion and conclusion
We conducted a comprehensive mixed-method analysis of the #IchBinHanna & #IchBin-
Reyhan movement, characterising its online discourse on Twitter and exploring contri-
buting factors to its success.
We conceptualise the movement as a connective action (Bennett & Segerberg, 2012),
driven by researchers connected through their collective identity as academic precariat
(Vatansever, 2023). Following the typology of collective and connective action networks
(Bennett & Segerberg, 2012), the #IchBinHanna & #IchBinReyhan movement is a self-
organising network with little to no organisational coordination of action. Formal organ-
isations such as unions, academic societies, and NGAWiss (Network for Good Academic
INFORMATION, COMMUNICATION & SOCIETY 19
Working Conditions) supported the movement, e.g., by organising related oine events,
but never played a central role in its organisation.
Since formal organisations with good connections to policymakers and the media are
usually associated with the ability to inuence policy, the movement’s substantial success
in inuencing policymakers and ultimately also legislation is surprising. An example
illustrating the successful transmission of the movement’s ideas to policy is the proposal
of the Anschlusszusage [promise of permanent employment] – a new legal instrument
supposed to alleviate the uncertainty about further employment at the postdoctoral
level. The instrument, first proposed by legal scholar and contributor to the movement
Simon Pschorr (Pschorr, 2023), was quickly taken up and publicly endorsed by other
voices within the movement. It features in the current draft amendment of the Wiss-
ZeitVG proposed by the German Federal Ministry of Education and Research (Bundes-
ministerium für Bildung und Forschung, 2023a) (though the way it is applied arguably
threatens to considerably worsen the situation of precariously employed researchers by
shortening the time they have to secure a permanent contract without requiring insti-
tutions to create more such positions). The movement was also successful in sustainably
attracting media attention (Bahr et al., 2023), considering that the situation only directly
aects a small proportion of the German population.
Our analyses of discussion topics, hashtags and participating groups reveal a number
of key factors that likely contributed to the movement’s rapid growth and wide reach.
Particularly at its onset, discussion subjects were diverse – from individual stories of pre-
carious working conditions to their connection with innovation, mental health, and pro-
posals for change – and highlighted intersections with other forms of discrimination.
Answering our first research question, the easy-to-personalise action frame oered by
#IchBinHanna & #IchBinReyhan dierentiated the movement from preceding move-
ments such as #FristIstFrust and #95vsWissZeitVG, allowing a diverse group of people
to connect with it. In addition, preceding movements and their associated hashtags lar-
gely co-existed, which may have split the attention of activists, policymakers, and the
media. In contrast, after the onset of #IchBinHanna & #IchBinReyhan, related hashtags
almost disappeared, rallying activists under a common theme which, in turn, made it
easier for policymakers and the media to respond and further amplifying the attention
that the movement received.
Building on this unified momentum, we observe a high diversity of groups participat-
ing in the movement. Answering our second research question, we observe that, while
the most directly aected group (postdocs) is clearly the most active, professors also
make substantial contributions – a group not aected by precarious working conditions.
This culminates in the creation of the new hashtag #ProfsFuerHanna, under which pro-
fessors rally to support precariously employed researchers in a form of performative soli-
darity (Morgan & Pulignano, 2020). In voicing their support, these professors emphasise
their shared superordinate identities as researchers in the German academic system to
connect with the precariously employed contributors to #IchBinHanna & #IchBinRey-
han (Selvanathan et al., 2020). The high variance of participation of professors between
topics, however, suggests selective and potentially strategic engagement of this powerful
group. Interestingly, the participation of students is generally low, even though they are
the largest group in the academic sector as a whole, and many students may well be
aected by the precarious working conditions that #IchBinHanna & #IchBinReyhan
20 A. WAGNE ET AL.
contributors decryst in the future.I The unification of the movement under a dominant
pair of hashtags enabled the dierent groups to remain united while allowing for the dis-
cussion of a diverse range of issues. Furthermore, the prominent involvement of pro-
fessors points to the high impact of intergroup solidarity from a powerful group. It is
likely that this intersectional coalition-building we observed expanded the scope of the
movement in a way that allowed it to grow rapidly. While the ensuing online discussion
falls outside our of observation period, media reactions to the #ProfsFuerHanna move-
ment suggest that the comparatively small number of professors involved succeeded in
generating a significant amount of media attention for the movement.
A limitation of our analysis of dierent groups participating in the discourse is the
large proportion of accounts (50.8%) that remain unclassified. However, these accounts
only represent 26.8% of tweets in our corpus. It is also worth noting that the unclassified
accounts are not an artefact of our computational classification approach, most of these
accounts could not be classified by human raters either. In the presentation of our results,
we merged the ‘unclassified’ category with accounts classified as institutions, media,
union representatives, political representatives, teachers and medical doctors because
of the low prevalence of tweets from each of these groups, individually. Analyses of
how tweets from these groups interacted with and potentially shaped the discussions
within the movement could be an interesting avenue for future research.
Considering our third research question, our analysis of the language used in the dis-
course reveals further factors that potentially contributed to the sustained relevance of
the movement. While some keywords such as befristet [temporary] were clearly associ-
ated with negative sentiment, the overall tone of the discussion remained supportive
and constructive. While negative emotions such as anger and outrage are associated
with increased engagement with social media content (Berger, 2011; Paletz et al.,
2023), there is also a growing body of evidence suggesting that such negative discourse
drives down engagement in the long run (Hickey et al., 2023; Ziegele & Jost, 2020).
The use of supportive and constructive language may provide an explanation for the abil-
ity of the movement to engage a substantial number of people over an extended period of
time. We also identified a topic criticising the movement, claiming that its main propo-
nents are social scientists and humanities scholars who do not ‘deserve’ good working
conditions as their contributions are ‘unimportant’. Deep reading of the tweets in this
topic reveals that the criticism remains on a superficial level, reiterating stereotypical
depictions of academic disciplines. It fails to engage in deeper reections on how research
and working cultures are shaped by disciplinary capacities and mandates to contribute to
societal development (Ploder et al., 2023). Within our investigation time frame, these cri-
tiques remained marginal on Twitter and did not gain traction.
Our linguistic analysis also provides insights into the movement’s success: many of the
most salient keywords are verbs, many of which indicate agency. Our qualitative analysis
of the topics also shows that, while many tweets discussed problems, a substantial num-
ber also proposed solutions. In addition, the communication largely remained civil and
constructive, throughout the observation time, reinforcing the impression of a unified
and worthy cause, with which policymakers and the media can interact without fear of
backlash. Another important factor is the direction of communication towards key actors
using @handle mentions that result in the mentioned account receiving a notification.
Many of the key accounts we identified are politicians, institutions, trade unions, and
INFORMATION, COMMUNICATION & SOCIETY 21
media representatives, who frequently reacted to these mentions. Furthermore, the high
prominence of keywords such as ‘structural’, ‘often’ and ‘frequently’ point towards the
identification of individual issues as a systemic problem, prompting the addressed actors
to deal with the issues on a systemic level.
The direct comparison of keyword analysis, topic modelling and deep reading also
highlights key dierences between the approaches: topic modelling excludes words
that are not generally found in the vocabulary of a language, like hashtags and mentions,
but that may be highly relevant to understand Twitter discourse. Keyword analysis cap-
tures the whole discourse, but requires the collection of a reference corpus. Furthermore,
keyword analysis lacks granularity in computational analysis of individual topics while
topic modelling struggles with describing a large corpus fully, due to computational com-
plexity, as well as challenges in selecting the right level of granularity. As both topic mod-
elling and keyword analysis fail to capture semantic detail like negations, deep reading is
necessary to interpret the findings. Together, the dierent streams of analysis allow for a
triangulation of results and the identification of robust patterns in the discourse.
We conclude that the #IchBinHanna & #IchBinReyhan movement has succeeded in
foregrounding systemic problems that have long existed in German academia. While
debates are still ongoing, we are hopeful that the issues and potential solutions high-
lighted by the movement will precipitate into changes to academic working conditions
in Germany.
Note
1. All translations provided in this article were done by the authors of the article.
Acknowledgements
We thank Christian Funk, David Adler, Maria Blöchl and Stefan Laser who were part of the #Ich-
BinHanna research collective, from which the present work emerged.
Author contributions
AW led the data analysis and contributed to writing the initial draft of the manuscript. ELF con-
tributed to the data analysis and contributed to writing the manuscript. FF contributed to writing
the manuscript. JL designed and supervised the research project, collected the data, contributed to
the data analysis, and contributed to writing the manuscript.
Disclosure statement
No potential conict of interest was reported by the author(s). ELF and JL were both active Twitter
users at the time of data collection and contributed to the #IchBinHanna & #IchBinReyhan
discourse.
Funding
JL was supported by the Marie Skłodowska-Curie grant number 101026507. FF was supported by a
Marietta Blau scholarship of the OEAD.
22 A. WAGNE ET AL.
Notes on contributors
Ahmadou Wagne is a doctoral researcher at the Department of Computer Science at the Vienna
University of Technology. His research focuses on preference elicitation in multi-domain recom-
mender systems with a conversational approach, especially in E-Commerce. Beyond this, his aca-
demic interests extend to the fields of natural-language-processing and computational social
science.
Elen Le Foll is a Post-doctoral researcher and lecturer in quantitative corpus linguistics at the
Department of Romance Studies & Data Center for the Humanities at the University of Cologne.
Her research interests include foreign language pedagogy, corpus-linguistic methods, learner
phraseology, media and data literacy, sociolinguistics, and the socio-cognitive processes involved
in second language acquisition and interpreting.
Florentine Frantz works as a consultant at the Technopolis oce in Vienna. She has a background
in Science and Technology Studies (STS) and Technical Physics, with extensive experience in
quantitative as well as qualitative research methods.
Jana Lasser is a Professor for Data Analysis at University of Graz and lead the research group of
Complex Social & Computational Systems at the interdisciplinary centre IDea_Lab. She researches
emergent phenomena in complex social systems, employing methods from machine learning, data
science, natural language processing and computational and statistical modelling [email: jana.
lasser@uni-graz.at].
Data and code availability statement
According to Twitter’s Terms of Service, we are not permitted to publish raw tweet texts. We do
however publish the IDs of all the tweets that were used in the presented analyses under accession
code 10.17605/OSF.IO/2M6Y8 (Wagne et al., 2024). Tweet IDs can be used to retrieve the corre-
sponding tweet texts as long as the tweets they reference have not been deleted. Since X’s (formerly
known as ‘Twitter’) API for retrieving tweets from tweet IDs is not functional at this point in time,
we will share tweet texts upon request for research purposes.
We created custom Python and R scripts to collect and analyse the data. All scripts are pub-
lished under accession code 10.17605/OSF.IO/2M6Y8 (Wagne et al., 2024). Since the academic
Twitter API to retrieve historical data ceased to function in April 2023, the code we used to retrieve
our data is not functional anymore.
References
Alfonzo, P. (2021). A topology of twitter tactics: Tracing the rhetorical dimensions and digital
labor of networked publics. Social Media + Society, 7(2), 20563051211025514. https://doi.org/
10.1177/20563051211025514
Alfonzo, P., & Foust, C. R. (2019). Campus activism in the digital age: An ecological chronology of
#Concernedstudent1950. Journal of Contemporary Rhetoric, 9(3/4), 87–111.
Bahr, A., Blume, C., Eichhorn, K., & Kubon, S. (2021). With #IchBinHanna, German academia
protests against a law that forces researchers out. Nature Human Behaviour, 5(9), 1114–1115,
Article 9. https://doi.org/10.1038/s41562-021-01178-6
Bahr, A., Eichhorn, K., & Kubon, S. (2020, November 30). 95vsWissZeitVG. 95vsWissZeitVG.
https://95vswisszeitvg.wordpress.com/
Bahr, A., Eichhorn, K., & Kubon, S. (2022). #IchBinHanna – Prekäre Wissenschaft in Deutschland.
Suhrkamp.
Bahr, A., Kubon, S., & Eichhorn, K. (2023). #IchBinHanna. #IchBinHanna. https://ichbinhanna.
wordpress.com/
INFORMATION, COMMUNICATION & SOCIETY 23
Bennett, W. L., & Segerberg, A. (2012). The logic of connective action: Digital media and the per-
sonalization of contentious politics. Information, Communication & Society, 15(5), 739–768.
https://doi.org/10.1080/1369118X.2012.670661
Benoit, K., Watanabe, K., Wang, H., Nulty, P., Obeng, A., Müller, S., & Matsuo, A. (2018).
Quanteda: An R package for the quantitative analysis of textual data. Journal of Open Source
Software, 3(30), 774. https://doi.org/10.21105/joss.00774
Berger, J. (2011). Arousal increases social transmission of information. Psychological Science, 22(7),
891–893. https://doi.org/10.1177/0956797611413294
Bird, S., Klein, E., & Edward, L. (2009). Natural language processing with Python: Analyzing text
with the natural language toolkit. O’Reilly Media, Inc.
Bössel, N., Kluge, A., Leising, D., Mischkowski, D., Phan, L. V., Schmitt, M., & Stahl, J. (2023).
Anreizsystem, Machtmissbrauch und Wissenschaftliches Fehlverhalten: Eine Analyse zum funk-
tionalen Zusammenhang zwischen strukturellen Bedingungen und unethischem Verhalten in
der Wissenschaft. Deutsche Gesellschaft für Psychologie. https://www.dgps.de/fileadmin/user_
upload/PDF/Berichte/Bericht_AMWF20230626.pdf
Bradler, S., & Roller, C. (2023). Befristung und gute wissenschaftliche Praxis: Biologie in unserer
Zeit, 53(1), 12–15. https://doi.org/10.11576/biuz-6206
Bundesministerium für Bildung und Forschung. (2021). Bundesbericht Wissenschaftlicher
Nachwuchs 2021. https://www.buwin.de/
Bundesministerium für Bildung und Forschung. (2023a). Entwurf eines Gesetzes zur Änderung des
Befristungsrechts für die Wissenschaft. Bundesministerium für Bildung und Forschung.
Bundesministerium für Bildung und Forschung. (2023b). Reform des WissZeitVG. https://www.
bmbf.de/SharedDocs/Downloads/de/2023/230317-wisszeitvg.pdf?__blob=publicationFile&v=1
Burrows, R. (2012). Living with the H-Index? Metric assemblages in the contemporary academy.
The Sociological Review, 60(2), 355–372. https://doi.org/10.1111/j.1467-954X.2012.02077.x
DeLuca, K. M., Lawson, S., & Sun, Y. (2012). Occupy Wall Street on the public screens of social
media: The many framings of the birth of a protest movement. Communication, Culture and
Critique, 5(4), 483–509. https://doi.org/10.1111/j.1753-9137.2012.01141.x
Deutscher Bundestag. (2021). Deutscher Bundestag Stenografischer Bericht 236. Sitzung. https://
dserver.bundestag.de/btp/19/19236.pdf#P.30621
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018). BERT: Pre-training of Deep
Bidirectional Transformers for Language Understanding. https://doi.org/10.48550/ARXIV.
1810.04805
Dirnagl, U. (2022). #Ichbinhannah and the fight for permanent jobs for postdocs. EMBO Reports,
23(3), e54623. https://doi.org/10.15252/embr.202254623
Djahangard, S. (2021, September 13). Berliner Hochschulgesetz: Schluss mit #IchbinHanna. Die Zeit.
https://www.zeit.de/campus/2021-09/berliner-hochschulgesetz-beschluss-arbeitsbedingungen-
wissenschaft-verbesserung-berufsperspektiven-universitaeten-debatte?utm_referrer=https%3A%2F
%2Fwww.google.com%2F
Entwurf Eines Gesetzes Zur Änderung Des Befristungsrechts Für Die Wissenschaft (2024). https://
www.bmbf.de/SharedDocs/Downloads/de/2024/2403_reg_entw_wisszeitvg.pdf?__blob=
publicationFile&v=1
Felt, U. (Ed.). (2009). Knowing and living in academic research: Convergences and heterogeneity in
research cultures in the European context. Inst. of Sociology of the Acad. of Sciences of the Czech
Republic.
Fochler, M., Felt, U., & Müller, R. (2016). Unsustainable growth, hyper-competition, and worth in
life science research: Narrowing evaluative repertoires in doctoral and postdoctoral scientists’
work and lives. Minerva, 54(2), 175–200. https://doi.org/10.1007/s11024-016-9292-y
Formanowicz, M., Roessel, J., Suitner, C., & Maass, A. (2017). Verbs as linguistic markers of
agency: The social side of grammar. European Journal of Social Psychology, 47(5), 566–579.
https://doi.org/10.1002/ejsp.2231
Fritz, F., Janotta, L., Marks, S., Prigge, J., & Schirmer, S. (2022). Lehr- und
Forschungsbedingungen: Wann kommt die Antwort der Fachgesellschaften auf
24 A. WAGNE ET AL.
#IchbinHanna? Debatte. Beiträge Zur Erwachsenenbildung, 4(1–2021), 97–101. https://doi.org/
10.3224/debatte.v4i1.05
Fritz, F., Kallenbach, T., & Kleverman, N. (2021). #Ichbinhanna – #WirSindNetzwerk. Soziale
Arbeit, 70(9), 322–328.
Gabrielatos, C. (2018). Keyness analysis: Nature, metrics and techniques. In C. Taylor, & A.
Marchi (Eds.), Corpus approaches to discourse (pp. 225–258). Routledge.
Gallagher, R. J., Stowell, E., Parker, A. G., & Foucault Welles, B. (2019). Reclaiming stigmatized
narratives: The networked disclosure landscape of #MeToo. Proceedings of the ACM on
Human-Computer Interaction, 3(CSCW), 96:1–96:30. https://doi.org/10.1145/3359198
German Ministry for Education and Research (Director). (2018). Wozu dient das
Wissenschaftszeitvertragsgesetz? https://web.archive.org/web/20210611145015/https://www.
bmbf.de/de/media-video-16944.html
Gewerkschaft Erziehung und Wissenschaft. (2021). #IchbinHanna – Per Hashtag gegen das
Wissenschaftszeitvertragsgesetz. https://www.gew.de/fileadmin/media/publikationen/hv/
Hochschule_und_Forschung/Broschueren_und_Ratgeber/IchbinHanna.pdf
Gibbs, P. (Ed.). (2015). Universities in the ux of time: An exploration of time and temporality in
university life. Routledge.
Goetze, D., Oetzel, L., & Süselbeck, J. (2023). #ResearchWonderland statt Brain Drain?
Internationale Stellungnahme zur geplanten Änderung des Wissenschaftszeitvertragsgesetzes
(WissZeitVG). Literaturkritik.De. https://literaturkritik.de/researchwonderland-statt-brain-
drain,29601.html
Grootendorst, M. (2022). BERTopic: Neural topic modeling with a class-based TF-IDF procedure.
https://doi.org/10.48550/ARXIV.2203.05794
Haug, K. (2022). Ich war Hanna. der Spiegel. https://www.spiegel.de/panorama/bildung/ein-jahr-
ichbinhanna-warum-junge-forschende-ins-ausland-oder-in-andere-jobs-ziehen-a-0713081c-
15ce-4b20-b319-79b9148e06f5
Hickey, D., Schmitz, M., Fessler, D., Smaldino, P., Muric, G., & Burghardt, K. (2023). No love
among haters: Negative interactions reduce hate community engagement. https://doi.org/10.
48550/ARXIV.2303.13641
Jackson, S. J., Bailey, M., & Welles, B. F. (2020). #Hashtagactivism: Networks of race and gender
justice. The MIT Press.
Jackson, S. J., & Foucault Welles, B. (2015). Hijacking #MYNYPD: Social media dissent and net-
worked counterpublics. Journal of Communication, 65(6), 932–952. https://doi.org/10.1111/
jcom.12185
Jakubíček, M., Kilgarri, A., Kovář, V., Rychlý, P., & Suchomel, V. (2013). The TenTen corpus
family. In Proceedings of the 7th International Corpus Linguistics Conference (pp. 125–
127). 23rd July 2013 to 26th July 2013.
Kreckel, R. (2008). Zwischen Promotion und Professur. Das wissenschaftliche Personal in
Deutschland im Vergleich mit Frankreich, Großbritannien, USA, Schweden, den Niederlanden,
Österreich und der Schweiz. Akademische Verlagsanstalt. https://www.hof.uni-halle.de/web/
dateien/Zwischen-Promotion-und-Professur.pdf.
Kreckel, R. (2016). Zur Lage des wissenschaftlichen Nachwuchses an Universitäten: Deutschland
im Vergleich mit Frankreich, England, den USA und Österreich. Beiträge Zur
Hochschulforschung, 38(1–2), 12–40.
Labour Rights Index 2022. (2022). WageIndicator Foundation and Centre for Labour Research.
https://labourrightsindex.org/lri-2022-documents/lri-2022-final-7-oct.pdf
Lasser, J. (2022). Twitter API v2 sampled stream. https://doi.org/10.17605/OSF.IO/DQX39
Lasser, J., Bultema, L., Jahn, A., Löer, M., Minneker, V., & van Scherpenberg, C. (2021). Power
abuse and anonymous accusations in academia – Perspectives from early career researchers and
recommendations for improvement. Beiträge zur Hochschulforschung, 43(1–2), 48–61.
Leinfellner, S., Thole, F., Simon, S., & Sehmer, J. (Eds.). (2023). Bedingungen der
Wissensproduktion: Qualifizierung, Selbstoptimierung und Prekarisierung in Wissenschaft und
Hochschule. Verlag Barbara Budrich.
INFORMATION, COMMUNICATION & SOCIETY 25
McInnes, L., & Healy, J. (2021). UMAP: Uniform Manifold Approximation and Projection for
Dimension Reduction (0.5) [Computer software]. https://umap-learn.readthedocs.io/en/latest/
McInnes, L., Healy, J., & Astels, S. (2017). Hdbscan: Hierarchical density based clustering. The
Journal of Open Source Software, 2(11), 205. https://doi.org/10.21105/joss.00205
Morgan, G., & Pulignano, V. (2020). Solidarity at work: Concepts, levels and challenges. Work,
Employment and Society, 34(1), 18–34. https://doi.org/10.1177/0950017019866626
Müller, R. (2012). On Becoming a “Distinguished” Scientist. Careers, Individuality and Collectivity
in Postdoctoral Researchers’ Accounts on Living and Working in the Life Sciences. | Lund
University. University of Vienna. https://www.lunduniversity.lu.se/lup/publication/ed500782-
d81b-4e88-b398-a3796cb69d12
Müller, R., & De Rijcke, S. (2017). Thinking with indicators. Exploring the epistemic impacts of
academic performance indicators in the life sciences. Research Evaluation, 26(4), 361–361.
https://doi.org/10.1093/reseval/rvx033
Nelson, L. K. (2020). Computational grounded theory: A methodological framework. Sociological
Methods & Research, 49(1), 3–42. https://doi.org/10.1177/0049124117729703
Netzwerk für Gute Arbeit in der Wissenschaft. (2019). Frist ist Frust. https://mittelbau.net/frist-ist-
frust/
Olson, M. (2003). The logic of collective action: Public goods and the theory of groups (21. printing).
Harvard Univ. Press.
Paletz, S. B. F., Johns, M. A., Murauskaite, E. E., Golonka, E. M., Pandža, N. B., Rytting, C. A.,
Buntain, C., & Ellis, D. (2023). Emotional content and sharing on Facebook: A theory cage
match. Science Advances, 9(39), eade9231. https://doi.org/10.1126/sciadv.ade9231
Papacharissi, Z. (2015). Aective publics: Sentiment, technology, and politics. Oxford University
Press.
Papacharissi, Z. (2016). Aective publics and structures of storytelling: Sentiment, events and
mediality. Information, Communication & Society, 19(3), 307–324. https://doi.org/10.1080/
1369118X.2015.1109697
Pfeer, J., Mooseder, A., Lasser, J., Hammer, L., Stritzel, O., & Garcia, D. (2023). This sample
seems to be good enough! assessing coverage and temporal reliability of Twitter’s academic
API. Proceedings of the International AAAI Conference on Web and Social Media, 17, 720–
729. https://doi.org/10.1609/icwsm.v17i1.22182
Ploder, M., Walker, D., Schibänker, H., Streicher, J., Bluemel, C., Knöchelmann, M., Müller, R.,
Sultan, A., & Simon, D. (2023). Wissenschaftskulturen in Deutschland. VolkswagenStiftung.
https://www.volkswagenstiftung.de/sites/default/files/documents/Studie_Wissenschaftskulture
n%20in%20Deutschland_web.pdf
ProfsfuerHanna. (2023). Nivellierung statt Novellierung: Kritik an der geplanten Reform des
WissZeitVG aus Sicht der Professorinnen und Professoren. https://docs.google.com/document/
d/1orgaZDC04YoH2PtDJBP5Ko1DgP-IWFIFPt8Ch3CDRME/preview
Pschorr, S. (2023, March 28). Ein Kompromiss, der seinen Namen verdient. Jan-Martin Wiarda
Blog. https://www.jmwiarda.de/2023/03/28/ein-kompromiss-der-seinen-namen-verdient/
Reimers, N., & Gurevych, I. (2019). Sentence-BERT: Sentence Embeddings using Siamese BERT-
Networks (arXiv:1908.10084). arXiv. https://doi.org/10.48550/arXiv.1908.10084
Richardson, L. (2007). Beautiful Soup [Computer software]. https://www.crummy.com/software/
BeautifulSoup/bs4/doc/
Rizzo, M., & Atzeni, M. (2020). Workers’ power in resisting precarity: Comparing transport
workers in Buenos Aires and Dar es Salaam. Work, Employment and Society, 34(6), 1114–
1130. https://doi.org/10.1177/0950017020928248
Selvanathan, H. P., Lickel, B., & Dasgupta, N. (2020). An integrative framework on the impact of
allies: How identity-based needs inuence intergroup solidarity and social movements.
European Journal of Social Psychology, 50(6), 1344–1361. https://doi.org/10.1002/ejsp.2697
Sigl, L. (2016). On the tacit governance of research by uncertainty: How early stage researchers
contribute to the governance of life science research. Science, Technology, & Human Values,
41(3), 347–374. https://doi.org/10.1177/0162243915599069
26 A. WAGNE ET AL.
Sigl, L. (2019). Subjectivity, governance, and changing conditions of knowledge production in the
life sciences. Subjectivity, 12(2), 117–136. https://doi.org/10.1057/s41286-019-00069-6
Sørensen, K. H., & Traweek, S. (2023). Questing excellence in academia: A tale of two universities.
Routledge.
Summers, E., Brigadir, I., Hames, S., Van Kemenade, H., Binkley, P., Tinafigueroa, Ruest, N.,
Walmir, Chudnov, D., Recrm, Celeste, Hause Lin, Chosak, A., Miles McCain, R., Milligan, I.,
Segerberg, A., Daniyal Shahrokhian, Walsh, M., Lausen, L., … Shawn. (2022). DocNow/twarc:
V2.10.4 (v2.10.4) [Computer software]. Zenodo. https://doi.org/10.5281/ZENODO.6503180
Theocharis, Y. (2015). The conceptualization of digitally networked participation. Social Media +
Society, 1(2), 2056305115610140. https://doi.org/10.1177/2056305115610140
Theocharis, Y., Lowe, W., van Deth, J. W., & García-Albacete, G. (2015). Using Twitter to mobilize
protest action: Online mobilization patterns and action repertoires in the Occupy Wall Street,
Indignados, and Aganaktismenoi movements. Information, Communication & Society, 18(2),
202–220. https://doi.org/10.1080/1369118X.2014.948035
Tilly, C. (2019). Social movements, 1768–2004 (1st ed.). Routledge.
Townsend, L., & Wallace, C. (2016). Social media research: A guide to ethics. University of
Aberdeen. https://web.archive.org/web/20220804092734/https://www.gla.ac.uk/media/Media_
487729_smxx.pdf
Vatansever, A. (2023). The making of the academic precariat: Labour activism and collective iden-
tity-formation among precarious researchers in Germany. Work, Employment and Society,
37(5), 1206–1225. https://doi.org/10.1177/09500170211069830
Wagne, A., Le Foll, E., Frantz, F., & Lasser, J. (2024). Reproduction materials for the article “Giving
the outrage a name” [dataset]. OSF. https://doi.org/10.17605/OSF.IO/2M6Y8
Wartena, C. (2023). HanTa—The Hanover Tagger [Python]. https://github.com/wartaal/HanTa
(Original work published 2019)
Wegner, A. (2020). Die Finanzierungs- und Beschäftigungssituation Promovierender. Deutsches
Zentrum für Hochschulforschung. https://www.dzhw.eu/pdf/pub_brief/dzhw_brief_04_2020.
pdf
Wilson, A. (2013). Embracing Bayes factors for key item analysis in corpus linguistics. In M.
Bieswanger, & A. Koll-Stobbe (Eds.), New approaches to the study of linguistic variability
(Vol. 4, pp. 3–11). Peter Lang. https://www.research.lancs.ac.uk/portal/en/publications/embra-
cing-bayes-factors-for-key-item-analysis-in-corpus-linguistics(3ecab40b-fd23-45ca-938e-
9b52756184.html
Wissenschaftlicher Dienst des Deutschen Bundestages. (2022). Zu befristeten Arbeitsverhältnissen
in der Wissenschaft und Innovation—Innovation durch Fluktuation (WD 8-3000-061/22).
Deutscher Bundestag.
Yang, G. (2016). Narrative agency in hashtag activism: The case of #BlackLivesMatter. Media and
Communication, 4(4), 13–17. https://doi.org/10.17645/mac.v4i4.692
Ziegele, M., & Jost, P. B. (2020). Not funny? The eects of factual versus sarcastic journalistic
responses to uncivil user comments. Communication Research, 47(6), 891–920. https://doi.
org/10.1177/0093650216671854
INFORMATION, COMMUNICATION & SOCIETY 27
