Women’s Histories in a Digital World: An Exploration of Digital Archives, Family History, and Domestic Violence in Early Twentieth-Century Australia

Abstract

In recent years, scholars have increasingly recognised the ways that colonialism, and related racism, embedded intergenerational trauma within families and communities. The role of domestic violence within families is widely accepted as important, but often treated separately. This article uses a case study from Western Australia, the life and death of Annie Grigo Dost, to explore the dynamics of both issues. Importantly, it also critiques the presentation of complex colonial family histories within a range of digital platforms, especially Ancestry.com. Such platforms obscure complex family dynamics, enforcing normative (often Westernised and highly gendered) digital frameworks for data, and consequently for stories about the past. This article offers an important critique of the ways that Ancestry.com in particular seems to actively sanitise family history, and the ways that they may be doing a disservice to their customers, who may want to acknowledge a more complex, critical family history.

1. Introduction

It is difficult to bring together the feel-good history marketed by the heritage sector with the messy reality of families. Katie Barclay has noted that families are often treated as a ‘key access point to heritage’, as if everyone has a family and that the associations are universally inclusive and positive (Barclay 2020). Anna Robinson-Sweet has similarly argued that Ancestry.com in particular presents a ‘fun’ and sanitised vision of the past, meaning that the company ‘effectively erases the violence at the root of many family trees’, especially for Americans with family connections to slavery (Robinson-Sweet 2021). This ties in with wider heritage industry practices, where Heritage is marketed and presented as a positive collective experience (Smith 2006). Family is a core tool used to link us with the past, to build that sense of positive and worthwhile connection.

As part of this special issue, I want to reflect on how family history is presented in digital platforms. I will consider why and how intimate family dynamics may be recorded (or not) and how that can shape our own research. The article explores how digital platforms like Ancestry.com and FindMyPast may sanitise family history, obscuring complex and potentially traumatic stories, especially for women and in settler colonial settings. It calls for more transparency in digital archiving platforms to prevent erasure of the real intricacies of families.

I use a specific case study to examine these issues: Anne Charlotte Grigo Dost (1880–1910) was born in Australia within a large and complicated family. Highly unusually, at the age of 16, she eloped with a man from colonial India (Baluchistan in modern-day Pakistan) who had migrated to the goldfields of Western Australia. Also unusually, her brothers murdered her husband, allegedly to protect her and themselves from his domestic abuse. She, in turn, went to colonial India to secure her and her children’s inheritance, where she was murdered by his family, allegedly in revenge.

Annie’s family life is far removed, then, from feel-good history. I learnt of Annie due to an administrative oddity: while constructing a database of all female applications for British naturalisation between 1901 and 1920 in Australia, I found a government file about her called ‘Mrs. Dost Mahomed Naturalization’ (Bright n.d.; NAA: A1 1919). I omitted her from that study because the file was not a naturalisation application. Uniquely, Annie’s 105-page file instead recorded newspaper cuttings and official discussions about the murders and the costs and complexities of returning the Australian-born children from colonial India to Australia after her murder (NAA: A1 1911, 1919). Why it was filed in this way is unclear.

Most contemporary information, like the ‘naturalisation’ file, is resolutely about their violent deaths. Headlines about both Dost’s and her murder emphasised him and his ‘othered’ status within Australia: ‘Mohammedan Killed’, ‘An Asiatic Killed’, or ‘Murder of an Afghan’ were all typical headlines (Barrier Miner 1909; Bendigo Advertiser 1909; The Daily Telegraph 1909). Contemporaries considered this as the ultimate tale of why white women should not marry outside of their race. In this regard, Annie’s life fits neatly within academic scholarship about interracial relationships, which has documented how local communities often responded by expressing fear and outrage on behalf of the women (Ray 2015). Especially in settler colonial societies, such relationships were frequently framed as leading to the woman’s degradation and even death, as well as resulting in wider racial and gender instability (see, for instance, Bagnall 2002; Cohen 2019). Annie herself had become an administrative headache for various colonial officials in India and Australia, and a warning for other young white girls.

My objective is to explore how we use digital records to do global history research like this, underpinned by feminist scholarship that has sought to reclaim women’s histories, and to read against the grain of colonial records (Stoler 2008; Abu-Lughod 1992). Annie’s life allows an exploration of the specific and the personal, in a local (or in this case, domestic) and a global context. In doing so, we can avoid using her to represent the experiences of an entire group (whether by race, gender, or even as a victim), something Mohanty has argued traditional women’s scholarship too often has done (Mohanty 2003). It was natural to use digitised records from the National Archives of Australia (NAA), Trove, Ancestry.com, FindMyPast, and an oral history testimony from a son, Ameer Dost, recorded in 1982 for the State Library in Western Australia (Dost 1982). Through these, I could conduct significant amounts of research. Due to her dramatic life and death, she is relatively well documented.

This, however, is also a critique of the limitations of commercial digital services, not just in terms of what is available, but also in terms of the opacity around how underlying platforms work. Annie’s family story, after all, fits poorly within the feel-good framework marketed by much of the family history sector. In this article, I will explore the histories of her and her family’s lives, but also reflect critically on the digitisation of records and the effects of digital platforms like Trove and Ancestry.com in presenting (or concealing) life histories like Annie’s.

To do this, I bring together digital and feminist scholarship with critical family history, which centres family histories within wider discussions of power relations, with enduring legacies in the present (Sleeter 2020). While critical family history has focused on the inherent violence of ‘the settler family’s home-making endeavour’, this study is both a more personal exploration of violence within families and an exploration of how wider issues around empire, race, and gender affected the records and stories told (Byrnes and Coleborne 2023).

Family trauma is long lived. As Andrew Milne explained in his own critical family history: ‘I did not spontaneously “happen”, but I am the sum of those that preceded me’ (Milne 2023). Private companies like Ancestry.com directly market themselves as offering the answer to who we are, through DNA and digital archives. In most countries, they even sponsor the popular ‘Who Do You Think You Are?’ TV show format, which centres this idea. Yet the current digital landscape’s tendency to strip out context can be a form of denial of the actual lived experiences of the past. Context matters: otherwise ‘race and racism may be “washed out” of stories’ about the past, along with other key factors like gender and class (Sleeter n.d.). Critical family history, rather than a bland ‘heritage’ model, can be painful but necessary for both academics and relatives. Digital platforms do not make this easy.

The distinction between academic and family historians is blurred here by the specific case study. I chose the case of Annie and her family for academic reasons. However, during my research, Ancestry.com identified within family tree ‘facts’ web pages that I have ‘a common DNA ancestor’ with each of Annie’s parents separately. What this means is unclear. They presumably do not have DNA from each person (they have been dead for approximately 100 years). Is this because one of their living descendants has been identified as a distant DNA relation? Ancestry.com does not make this clear. After exploring their relatives, this label is also attached to Henry Grigo, the brother who killed Annie’s husband, and to one of Henry’s children, but not to Annie herself or other siblings. It is likely Henry’s child or grandchild who is the actual DNA link, but this remains guesswork. I will discuss many examples of similar lack of clarity within Ancestry.com in this article. Regardless, this was a novel experience for me: my academic work was also now a family history, of sorts.

2. Digital Women’s History

Digitisation has long been associated with democratising access (you no longer have to physically travel) and offers significant social history potential (Inwood and Ross 2016). Family history has been one of the predominant drivers of digitisation so far. Such history is consciously marketed as a worthy task, recovering lost, often working class and female voices in the finest traditions of social history (Robb 2024). A recent FindMyPast article told readers specifically how to find female ancestors, explained what social history was, and offered their service as the ideal way to uncover women’s hidden but important past (FindMyPast 2024). Companies like FindMyPast and Ancestry.com specifically target women in marketing and feature real and fictional women’s stories in advertisements (Ancestry UK 2013; Ancestry Canada 2019; Ancestry.com 2020). In reality, digital tools seem to make family history easier but can also obscure women’s lives. Most obviously, by focusing on families, they accentuate traditional female roles; other problems can be less obvious.

At issue are content and access. What does access to a digital archive actually provide? It often offers a false sense of completeness. This is especially true when using paid-for platforms like Ancestry.com or FindMyPast to navigate several countries at once; there are considerable differences between which national records are available, based on which records have been collected and microfilmed by partnering libraries and archives. Issues of copyright and enforced standardisation accentuate problems further. The algorithms used, licensing agreements, and access costs are the opposite of democratising. In the words of Jerome de Groot, private companies like Ancestry.com ‘have replaced the academy as gatekeepers and controllers of access to knowledge’ (De Groot 2020), but often do not acknowledge this role. Similarly, A. Gomez, reflecting on working class British history, has noted that ‘the old politics of who deserved to be historicized has become the new politics of who deserves to be digitized’ (Gomez 2019). This choosing of what to digitise influences what is available. In Annie’s case, digital material is readily available in Australia, while no digital records exist about her periods of life in colonial India, the son she had (and left) there, nor about her murder there. Such inclusions and absences can significantly skew the view users have of the past. Her Australian life is relatively well documented. Her life in colonial India is completely absent, reflecting ongoing postcolonial digital inequality.

Of course, simply having a lot of data is not useful. There must be some way to classify and store data, to organise them to make them searchable; ‘findability’ is crucial for us to be able to use what David Canning calls the ‘digital heap’ (Canning 2023; Zaagsma 2023). This requires standardisation, which can mean that some ‘data that is nonstandard in some sense registers as visible only when made to give up its variation and conform to the norm’ (Bhattacharyya 2022). The largest company doing this is Ancestry.com, with over 60 billion documents digitised, according to their website. Their services are available in 128 countries. Their website informs the reader that ‘there’s endless opportunities to find family connections and better understand the different regions that make up your estimate’ (Ancestry.com n.d.a, n.d.b, n.d.c, n.d.d). Originally based on the archives of the Mormon church in the United States, Ancestry.com is now owned by The Blackstone Group and had an annual revenue in 2022 of USD 1 billion (Robinson-Sweet). Scholars have increasingly noted that the company is not simply a neutral repository; the company chooses sources and builds both a new archive and an infrastructure to ‘share’ and ‘preserve’ data while seeking to ‘empower’ its paid users (Ancestry.co.uk ‘Legal Overview’, quoted in De Groot 2020).

The processes and priorities shaping what is digitised and how that metadata is organised means that poorly understood gaps arise. One scholar confidently predicted in 2009 that ‘Unless the copy is a bad one, no actual information is likely to be lost’ (Davison 2009). This is not the case. We rarely consider the underlying structures of a database. Indeed, even seasoned Digital Historians building their own database may simply throw together an Excel spreadsheet, to get to the ‘important’ analysis as quickly as possible (Inwood and Ross 2016). While historians may omit sufficient planning and reflection due to time and resource constraints, paid-for family history companies have a commercial interest in hiding any suggestion of incompleteness or enforced standardisation. The norms for family history have largely been established by the original Ancestry customer: white Americans. Users are all given little guidance to help them understand that Ancestry.com is not a historical source itself. It is a digital platform. Some information is free, but the ‘actual’ historic documents are available only to paying members who must click and view separate scanned documents. Some documents require a premium membership. However, to get to that original document, one has had to execute a word search, and that raises many potential spaces for error or for imposing standardisation.

Even in the best digitisation projects, the potential erasure of non-standard data is increasingly recognised as a challenge, an especial concern for historic data (Zaagsma 2023; Risam 2015; Coburn 2021). My omission of Annie from my own database is an example: her file did not fit, so was not included. None of the records (censuses, business lists, electoral registers, or migration files) usually available on family history sites were ever neutral organisers of raw data:. Women, the poor, and minorities were less likely to be named, while the illiterate may have had incorrect data recorded. Women and migrants changed their names. Surnames are often assumed to be paternal, although this practice is not universal. Documents will often be under the name of the head of the household, who was almost always male. The order of first and surname depends on culture. It is easy for women to disappear, rendered invisible, either on the original document or through search functions.

This can mean that structural problems in the past are replicated uncritically, such as assumptions of whiteness. In Australia’s national records about migration and naturalisation, officials and cataloguers made no racial identification except in a few cases (usually connected to China or ‘Syria’); otherwise, the assumption of whiteness was implicit and widespread. In the words of Ruth Frenkenberg, such an ‘unmarked, unnamed status’ for whiteness ‘is itself an effect of its dominance’ (Frankenberg 1993). There is little incentive for private companies to explore the omissions or structural problems within their records. Such assumptions can easily become embedded in newer digital platforms. Robinson-Sweet has noted how Ancestry.com treats customers in the U.S. as white unless there is explicit evidence to state otherwise. However, if identified as African American, white ancestry appears to only be identified as possible if there is a specific genetic match. Otherwise Ancestry.com assumes African Americans and white Americans do not share ancestors (2021).

There is an additional problem with the status of family history within academia. Genealogy research has often suffered from a widespread academic snobbery against amateur historians who have not been formally trained in how to research and interpret the past. As readers of this journal will know, academic History (and wider Higher Education) traditionally viewed genealogy as inferior to ‘important’ history (Sleeter 2020). Women’s lives, especially when focused on domestic or individual histories, have also often been rendered ‘invisible’ within academia, considered marginal, passive companions to the men who actively made history (Kamp 2022). This can have unintended consequences. In Australia, naturalisation paperwork has been organised by individual in the National Archives of Australia (NAA). The NAA’s guidebook describes these files as ‘primarily of interest to family historians’, a description that marginalises them from broader academic histories of migration and citizenship.

In addition to questions about academic rigour, there are usually different levels of prestige attached to different kinds of history. Local or family history often suffer from a belief that the small scale of such projects makes them unimportant in explaining wider processes of change over time. For academics, family history usually must be linked to wider, ‘important’ issues to justify what we do. Families are common. Academic history needs a wider significance, even within Social History.

This usually means focusing on historically underrepresented voices such as colonised or female subjects, reflected in the fact that this is the second prominent special issue within Genealogy about colonialism and family history. Even those approaches can go un-rewarded: Family history research projects can seem less prestigious precisely because their small scale means that external funding is not always needed. Prestige, promotion, and jobs are often strongly linked to the ability to secure funding for research. If you do research that does not require funding, then that makes you, and your research, less prestigious, regardless of the actual quality of work produced.

We can also see the impact of these issues in how genealogy companies market themselves. FindMyPast, for instance, launched a mirror service in 2022 of a digital archive called The Social History Archive. This is seemingly the same as their long-standing FindMyPast service but with a different name, one aimed specifically at academic audiences within schools and universities. University libraries may balk at funding FindMyPast, but not The Social History Archive, which sounds much more academic, even if the content is identical.

This makes for a noticeable difference between how academics, especially historians, design or use digital platforms, with a focus on the ‘bigger picture’, and how private companies design their digital platforms, focused on the ‘individual’ ancestor. These different approaches to family history can have important consequences for how we see and interpret the data available. In the Australian case, the naturalisation files are digitised but require a researcher to know the exact spelling of the name to find the relevant file. In Annie’s case, the name given is not obviously hers, and as it is not even clear why she was filed in the naturalisation papers, as it seems unlikely anyone would even look for her file.

The NAA files have also not been incorporated into Ancestry.com’s or FindMyPast’s records as they are already freely available. This, however, can make them invisible to family researchers, as there is no link or signpost connecting the different digital collections to each other. For historians, the sheer scale of migration and naturalisation records can be off-putting too; my own cataloguing of files took over five years. There is a reason why these files have been little used to date (although Kate Bagnall has done much to champion their potential for examining Chinese Australian family histories). Marking these records as for family history can render them invisible to academics and makes people without descendants also invisible and, by extension, unimportant.

More broadly, private companies take documents like ship lists and censuses and parcel them up within new search functions that tend to only allow the user to research individuals. This is a significant change from the original design and function of some of the most common data, such as censuses, but this process is rarely explored. If I want to track a particular individual, this can be a great function. However, in its original format and intention, scholars could easily construct some metadata analysis of a census district’s population, such as where people were born or what occupations they had, or how many lived in a house on average. Most of the first digital history projects did this type of work (Crymble 2021). The original intention and design are completely stripped by focusing now on individual ancestors. Lists are converted from written names along a street into digital text libraries of names, often using human and increasingly A.I. to ‘read’ and write names. The original organisation by location is stripped away from the search function to allow access only to the individual.

This is significant in terms of altering the ways we may access or understand the data. Recently, Living with Machines, ‘the largest digital humanities project ever funded in the UK’, brought together British historians and digital experts from the Alan Turing Institute to examine nineteenth century mechanisation through already digitised sources (Ahnert et al. 2023). They had prior permission to access complete census returns, which had been digitised by FindMyPast. The data had even been cleaned and coded by a previous publicly funded project so no problems with access were expected. However, a clause allows digital data owners to ‘safe-guard their data’, and in this case FindMyPast chose to invoke this, as ‘census data forms the central plank of FindMyPast’s subscription service to (mainly) family historians’. The project was instead offered an anonymised version of the census data, with columns removed that would have identified names and addresses. As the project organisers noted, the inability to see complete data made analysis impossible. Furthermore, ‘there is clearly an ethical concern that users must pay a fee to access resources already funded from taxpayers’ money.’ The academics had to apply for a special licence to access the full data, but that ‘proved difficult and drawn out’, taking months just to complete the paperwork involved, and many more months to actually receive the data, leaving very little time in the project for analysis (Ahnert et al. 2023). Again, these were national census data held by TNA, digitised by FindMyPast, cleaned by public funding, and ostensibly open to the British public for research purposes, but the complexities of accessing the metadata would be insurmountable for most researchers.

Access is not the only issue. Each layer of administration and alteration adds a layer of potential error. For example, Elizabeth Ann Stretcer (a woman naturalised in early twentieth century Australia), was wrongly identified as a man on Ancestry.com’s website’s listing from the 1937 electoral roll, despite the original document correctly labeling her as female. This is clearly a transcription error (Ancestry.com 2010). It is impossible to know how often such errors occur, let alone their impact. Errors can be corrected by a family member. Stretcer had no obvious relations, and it is not clear what steps (if any) Ancestry.com takes to check data. Because private companies want to advertise how well they work, and protect the underlying digital infrastructure, the processes of transcription and formatting are deliberately opaque. Importantly, who corrects the data of the un-familied?

It is also not always clear what data are put together by machine learning, and what is from people using the services. Ancestry.com, for instance, has a LifeStory function for any individual, appearing to be a combination of family historians’ family trees and automatically generated data, using the same language and layout for everyone. There are core dates associated with a person displayed, usually birth, death, and marriage, occasionally (but not always) linked to original documents. As in the Stretcer case, looking at scanned documents closely might identify an error. However, because it is not always clear where the data originated, or how it has been processed to fit into their search engine, it can be difficult for users to find or determine accuracy.

Omissions are another difficulty. In the immortal words of Donald Rumsfeld, there are unknown unknowns throughout the process. You enter in a name, and hope that the original document recorded their information correctly, and hope that the digitisation process ‘read’ and wrote the name correctly into the digital archive. If you are lucky, you get a match. Perhaps your ancestor had a common name. Perhaps it was spelled incorrectly on the original document. Perhaps their writing was messy, or the copies made onto microfilm were not clear. Perhaps the existence of a ‘head of household’ caused a person in that same residence to have an assumed surname or occupation that may not be correct. Perhaps they were never recorded on the census. Even family tree formatting on most websites creates a standardised structure and presentation, a highly heteronormative list that emphasises names and dates and marriage and procreation facts over any wider context. At every turn, it can be difficult to bring together context and individual, even if we wish to break down such barriers. Should we uncover ‘accurate’ digital records about an individual, there is a significant sleight of hand about what these records can reveal about individual lives.

3. Case Study: Anne Charlotte Grigo Dost

Let us look at the digital records about Annie. Here, I will explore records from the time, mainly newspaper coverage available for free through Trove, as well as the interview with her son and her ‘naturalisation’ file. The next section will compare this with how her life is depicted on Ancestry.com.

Annie, shown in Figure 1, was an Australian-born woman of Prussian–Danish heritage; even before her unusual elopement, this made her an outlier in a colony where over 95% of settlers at the time were British-born or had British parents. What research exists has primarily focused on her husband, a member of the iconic ‘Afghan’ cameleers in Western Australia. ‘Afghans’ was the term given to a small group of migrants from Northwestern India and surrounding areas, most of whom were Muslim, who acted as goods suppliers across Western Australia in the later nineteenth century. They imported and controlled extensive camel transport until lower gold prices and the creation of water and railway infrastructure in the early twentieth century made them obsolete. The roughly 4000 cameleers who arrived between 1870 and 1920 have received significant recent attention, especially Dost, who was economically important and had such an unusual marriage and death. A recent ABC radio piece profiled Dost as an example of the cameleers and their economic contributions (ABC 2024; Facebook Family History 2019). This was itself connected to a regional museum exhibition on cameleers, which in turn drew heavily on the interview conducted with the Dosts’ son in the 1980s (Rajkowski 1987; Willis 1992; Owen 2002; Stevens 2002; Kabir 2004; Sheardon 2008; Westrip and Holroyde 2010; ‘Afghans’ 2013; Kalgoorlie History 2017; Butta 2022; Day n.d.).

Mahomet Dost even has his own Wikipedia page with three main focuses: the ‘camel business’, of which he is presented as a leading figure’, their unusual ‘Australian family’, and their ‘Deaths’, with the violence clearly acknowledged and discussed (Wikipedia n.d.). However, a different Dost Mahomet is clearly confused with him, as there is also reference to the Burke and Wills expedition; their expedition was led by a cameleer of a similar name, but that Dost died in 1880 (Jones and Kenny 2010 explain the difference between the two men). It seems likely that this confusion is exacerbated by the lingering confusion about the order of his name in available records, whether he was Mahomet Dost or Dost Mahomet. Their children certainly considered Dost the surname, so I have also adopted that name order.

Annie is largely invisible from records until 1909, when her two brothers were charged with murdering Dost and tried at Broome, Western Australia. Despite extensive coverage, and Annie’s place as witness and victim of violence, she did not give evidence. Apparently out of a sense of chivalry, when she was ‘called as the last witness for the defence, the jury informed the Court that they were prepared to acquit the accused’ (her brothers) without hearing her evidence (Mudgee Guardian and North-Western Representative 1909). Even in this widely covered event, when protecting her was the reason given for the murder, she is rarely discussed.

According to George Worth Fry, coxswain of police at Port Hedland, who arrived just after the fight, he was told by Annie’s brother William immediately afterwards that the struggle occurred one evening when William had just arrived and Annie was making a cup of tea: ‘‘just as she was doing so Dost came out and spoke to Mrs. Dost in Afghan which he could not understand, and immediately kicked her in the stomach; Dost then put his left hand on her shoulder, and attempted to strike her with his right’. William stepped in ‘and told him not to hit his sister in his presence’, at which point Dost, an experienced and much celebrated wrestler, ‘grabbed’ him ‘by the throat’ and threw him, then strangled him. In the ensuing fight, both hit their heads, but Dost was on top, until the youngest brother, Harry, arrived in the room and hit Dost on the head. Only then did Dost stop the attack. Dost died of head wound hours later. The policeman went on to explain that: ‘Dost was a very powerful man; if on equal terms Willie Grigo would have no chance with him’ (Broome Chronicle and Nor’West Advertiser 1909a).

When William, the elder accused brother, gave his testimony on the third day, the focus was mainly on his relationship with Dost, not Annie. ‘Willie’ had known Dost since he was about 12, and Dost ‘was always very gentlemanly and kind’. William ‘at first had a strong prejudice against him’, apparently because of his race, ‘but his gentlemanly bearing and kindness soon removed this’. Several years later, when around 20 years old, Dost had brought him into the business of cameleering. As shopkeepers increasingly refused to sell to Dost directly as he was ‘an Asiatic’, Dost asked William to get licenses to carry out trade. William agreed so the business was legally in William’s name, but he was really working under Dost in the business. According to William, everything was amicable until the financial strain (suffered by all cameleers around 1909) seemed to make Dost prone to temper. The financial strain was because competition and reduced cameleer income had led to a recent camel strike. According to William, ‘Dost was at the head of that strike’ and began to always carry a rifle or stick for protection, as some cameleers were violently assaulted. William thought Dost had become more prone to lose his temper, given all of this. William concluded that ‘Dost was a dangerous man when put out’ and he had expected that ‘Dost would have killed him’ if not for his brother’s intervention (Broome Chronicle and Nor’West Advertiser 1909b).

When Harry gave evidence, he too emphasised that he had ‘always been friendly with Dost’ but that ‘on the evening of the 6th [the day before his death] Dost did not seem in a good humor’, and William told him it was because Dost had ‘money difficulties’. That evening was ‘the first time’ Harry had ‘heard of him ill-treating my sister’, witnessing her ‘trembling and frightened’. Dost had even thrown ‘his boot at her’ before Harry went to bed at their house. The next morning, he awoke to the fighting and intervened to protect William, not to kill Dost (Broome Chronicle and Nor’West Advertiser 1909b).

Annie was not the only silent witness to the death of her husband. Newspapers reported that the ‘chief’ witnesses for the prosecution were Stapleton (his identity is unknown, and no first name is given) and Omrah (probably another Afghan cameleer). Despite suggesting that they were at the property during the fight, it is unclear how much they witnessed because their testimony was not recorded by the press. Readers were merely told that both were ‘shaken by cross-examination’ and vague on details (Hedland Advocate 1909). On the final day, dramatically just before Annie’s testimony, the ‘not guilty’ verdict was read out. The judge, Mr. Wood, Commissioner of the Supreme Court, ‘complimented them on the attention they had given this most protracted case’ and ‘quite concurred with the verdict’ (Broome Chronicle and Nor’West Advertiser 1909a).

Historians of empire are long used to such an erasure of the voices and experiences of marginal groups in colonial settings. Annie did not speak in a male effort to protect her, but it also rendered her voiceless; her experience did not get recorded. Historians understand that the silencing of her and the two ‘chief’ witnesses reflects racist and sexist thinking at the time. Ancestry.com and FindMyPast also silence these people, because they do not even acknowledge that the murder happened, as I will discuss in the next section.

While contemporaries did not give us their testimony, we know from later sources, including their son’s oral history, that the local Afghan population and Ameer believed that this was murder, not a domestic dispute. Ameer claimed that the brothers murdered Dost while drunk because they were racist and because it allowed them to claim the business as their own (Dost 1982). However, it is worth remembering that Ameer’s telling was second hand and contained clear inaccuracies. Afghans, however, had reason to be suspicious of official narratives. A few years before, two Afghan cameleers (Jehan and Noor Mahoment) were murdered by two white competitors who were then found not guilty by a clearly racist jury. Other nonfatal racial attacks occurred regularly and went unpunished (Willis 1992). In this environment, mistrust seems reasonable.

Regardless of whether this was murder or self-defense, two things are clear: domestic violence did occur, and Annie was subsequently deliberately lured by Dost’s brother to India in order to murder her. Dost’s brother, acting as executor, convinced Annie that the only way she could claim her husband’s alleged wealth was to travel with the children to colonial India, a place she had lived for at least two years when she had first run away with Dost when 15 or 16 (and where the eldest child was born and remained). Annie was accused of supporting her brothers’ self-defense arguments, allowing them to escape justice. According to Ameer: ‘my mother got the blame for this… The Indians are an eye for an eye and a tooth for a tooth you know, it’s in their tradition isn’t it? So naturally they would think that they would murder her for this, because she got them [her brothers] out of gaol, you know, being hung’ (Dost 1982).

Such a view, that Annie’s death was to be expected, was one manifestation of Ameer internalising the racism of the period, mirroring the coverage at the time. The Argus reported it with the heading: ‘MARRIED AN ASIATIC. AUSTRALIAN WOMAN’S FATE. ELEVEN KNIFE WOUNDS’, as if this was the logical sequence of events (Argus 1910). Perth’s Sunday Times provided extensive coverage with the subheading: Dost Mahomet—A Native of Beluchistan—Subject to Paroxysms of Brutal Ferocity—Some Reminiscences of Coolgardie—A Courtship, a Family Feud and an Abduction’, clearly based on direct correspondence with William (Sunday Times 1910b). The Sydney Morning Herald pointedly headlined their coverage: ‘MURDERED BY AN AFGHAN. AUSTRALIAN WOMAN’S CAREER’ (Sydney Morning Herald 1910). Coverage generally emphasised her husband’s violent death and the role of Dost’s own brother in luring her to India, which was generally described as an ‘Afghan blood feud’ (The Daily Telegraph 1910; Herald 1910; Sunday Times 1910a; The West Australian 1910; Western Mail 1910; Zeehan and Dundas Herald 1910; Darling Downs Gazette 1910; Maitland Weekly Mercury 1910; Coolgardie Miner 1911; The Daily Telegraph 1911).

Australian press coverage firmly situated events outside societal norms in Australia. The Truth in Perth described Karachi as ‘far removed from the ordinary confines of civilisation’. The relationship was described as ‘the obstinate infatuation of a young woman for an Afghan suitor’, while her murder was ’a gruesome corroboration of the arguments used to induce that young woman to break off all connection with Afghans.’ Even Annie’s mother was implicitly criticised for renting accommodation to ‘Afghan residents’ in Perth at one time. The paper concluded:

no matter what may be the actual facts of the murder they furnish another vivid set of circumstances which should impress upon the people of the Commonwealth the awful danger of Afghan alliances. More particularly so when the unfortunate white wife enters the confines of those Asiatic countries where turbulent men are a law unto themselves when women are concerned.

(Truth 1910)

The implication was clearly that domestic violence and familiar murder were not part of typical (white) Australian life. While never explicitly discussed, the underlying argument was that white men did not do domestic violence. Such a myth was encouraged by the reporting of Annie’s story. In contrast, significant scholarly research has shown how colonial frontiers were especially violent spaces. Australia was no exception (Moore 1998). The celebration of violent masculinity, especially when accompanied by the hardships and isolation of colonial settings, often led to domestic abuse (Saunders 1984). This could be exacerbated if one was subjected to colonial racism and other forms of violence, as Dost was (Curthoys 2020).

We do not know how often Dost was violent. At the trial for Dost’s murder, her youngest brother thought it a new thing due to economic strains. Even the sympathetic remembrances of their son in the 1980s mentioned it: Dost ‘would lose his block like all our family, and my sisters and myself and my sons even too, are the same too, all of us’, he recalled, but then they would calm down quickly (Dost 1982). Dost’s prowess as a ‘wrestler’ was also clearly a part of his reputation across Western Australia, a reputation still celebrated by Ameer in the 1980s and clearly a factor in his own choice of career as a boxer. Ameer also suggested that the Grigo family was also violent, but such violence was less visible; they were never subjected to public scrutiny in the same way. From various family accounts, and other retellings, the Grigo–Dost family life is remembered as loving but ‘often not harmonious. The brothers were heavy drinkers, sometimes violent and not always respectful of Muslim practices’, while her husband was a keen wrestler and ‘had a reputation for quick temper and there are reports of physical violence in the home’ (Facebook Family History 2015).

Contemporary portrayals of domestic violence as uniquely ‘Afghan’ were possible because domestic violence was usually less visible than other types of colonial violence. Some Australian scholars have suggested that there has been a longstanding legal effort in Australia to actually make domestic violence invisible (Ailwood et al. 2012). We only know of this violence because it ended in murder; most domestic violence is hidden away.

In Annie’s case, while her life was clearly more than the two murders, the violence is unavoidable, affecting both her and her children. Born in 1904, Ameer Dost, known sometimes as Arthur or Dusty Doust, was the youngest child of Annie and Dost, and clearly had his own complex interaction with his parents’ lives and deaths. While too young to always remember accurately, his oral history provides striking details about their lives and family dynamics, with Annie prominently discussed. His parents lived on a sheep station near Port Hedland, ‘very happily too they were there I believe’, despite the occasional domestic violence and difficult relationship with Annie’s family, who liked to gamble and drink, while his father was usually teetotal as a practicing Muslim. He remembered his parents’ physical beauty, which he thought was why Annie put up with Dost being ‘quick tempered’ (Dost 1982).

In fact, we know quite a bit of detail about other things too. We can see that she came from a close, if volatile, family. We know that she and her sister loved to go to the races; Ameer thought that his parents had met at the races, although elsewhere it was stated that they met while she worked at the bakery managed (but not owned) by her father. Her father was frequently drunk and clearly absent often. He disowned her when she ran off with Dost at the age of 15 or 16 to colonial India. When they returned to Western Australia, Dost lavished the family with money and Annie with jewelry. Her father was the only family member to never accept Dost as a husband, due to strident and vocal racism. The rest of her family eventually moved to be near her and worked for Dost in one capacity or another. The brothers were cameleers while her sister received a loan to buy a local hotel. Dost was away for most of the time with his cameleer business, with Annie left on their farm near Port Hedland, where she tended to goats and her children, with her siblings often stopping by, as on the day of Dost’s death. She had many expensive rings, bought for her as presents by Dost, stolen when she was murdered, and she liked dressing up. She also once chopped off her own finger after being bitten by a poisonous snake when home alone looking after the children. She was reported to be very calm about it, just as she was reported to be calm about returning to India, even while suspecting it was a trap (Dost 1982; NAA files; newspaper reports). We know she verbally told several people, including police and a magistrate, when about to go to India, that she wanted her sister to be the guardian of the children if anything happened to her. The children remembered her and Dost with considerable love.

It is also clear that Annie’s children were deeply traumatised. They had lived with love and domestic violence, with the murder of their father by their uncles, witnessed their mother’s murder, probably by cousins (Ameer offers a harrowing account). They had believed their father a rich prince, yet when they were finally returned to Australia, they were penniless, put in orphanages, and their father’s wealth had vanished. Some blamed the ‘Indian’ relatives, especially Dost’s brother as executor and the man who convinced Annie to go to India. Some blamed the Grigo brothers for cheating the children out of their inheritance. The estate was considered intestate in 1912 and held ‘in trust’ by the government, valued at £800, all in camels (The West Australian 1912). It is possible that, given the decline of the camel industry, the value was overstated, that the government retained the money to cover expenses, or that the Grigos kept the money (as Ameer believed). Whatever money was left, it is not clear what happened to it.

It is also not clear why the children went to orphanages rather than to relatives. Ameer describes a few attempts by Indian relatives to convince him to run away back to India, but he was too scared of them, and the memories of his mother’s murder, to participate. While often the press was obsessed with returning the ‘Australian’ children to Australia, in practice, the children were described as, at best, pitiable, and at worst, semi-savage ‘Indians’ who hardly spoke English. For whatever reason, the children were placed immediately upon their return to Australia in various orphanages, where further trauma and abuse occurred, including racist physical attacks (Dost 1982). It seems safe to say that intergenerational trauma occurred.

In short, Annie lived an eventful and complex life, as did the children born in Australia. The child born and left near Lahore has proved almost invisible in digital records, although some family connections remain. The family dynamics are clearly complex and continue to interest family members, active on Ancestry.com, Facebook, and Wikipedia.

4. Ancestry.com’s Family Tree: Facts and LifeStories

Ancestry.com presents itself as the premier source for researching family stories. This should be a relatively easy case, given that Annie and Dost are well documented. Trove ensures that the sensational newspaper coverage in Australia is freely accessible (if you can be creative about name searches), while her ‘naturalisation’ file is available as a pdf through the NAA catalogue. There are also the family histories on Facebook, Wikipedia, and Ameer’s oral history.

None of these records, however, are identified or used by Ancestry.com. Indeed, the story of their lives that appears on Ancestry.com is almost insipid in its blandness. Scholars have become increasingly aware in recent years of ‘archival silences’: this can be both silences in the original materials and in what was deemed ‘worth’ keeping (Carter 2006; Cook 2011; Thomas et al. 2017). Annie herself was silenced in her lifetime, and now again on Ancestry.com, which presents her life and relationships in a way that erases violence. Many of the problems with digitisation are also evident.

One problem is about names. Her father is listed as Carl Griego, from Prussia (now Russia), while his children are all Grigo, although Annie’s birth certificate appears to say Grego (Ancestry.co.uk n.d.c). Her mother was Maren “Marie’ Hansen, from Horsens, Denmark. Annie was born in 1880 in Clermont, Queensland (some Ancestry.com sources incorrectly say 1884) and died in Karachi in 1910. As is typical on Ancestry.com, the viewer is presented with a ‘Family tree: Facts’ web page. For her and her parents, the only personalised information is that, when Annie was four, her father had a warrant issued ‘by the Water Police Bench… charged with unlawfully deserting his wife, Mary, leaving her without means of support’ (Ancestry.co.uk n.d.a, n.d.b, n.d.c). On this occasion, he clearly returned to the family, as Annie had another four siblings born after this date, including one within the year. This is the only hint on Ancestry.com of family disharmony.

Annie’s husband is listed as Mahomet Dost consistently in the timeline, but elsewhere in related documents is sometimes called Doust Mahomet; Mahomet is spelled sometimes as Mahomed. Sometimes she is Annie Grigo Dost, sometimes Annie Dost or Annie Mahomet. To further confuse things, the anglicised spelling of Doust was adopted by the children after their return to Australia from colonial India, in order to obscure their link with the scandal (and perhaps their ‘Indian’ heritage). Ameer recalled: ‘I did not want anybody to know of the murder so I agreed with all my sisters that we would change our names’ to Doust (Dost 1982). (It is worth noting that the family later reclaimed the original spelling).

None of this is evident on Ancestry.com. Along with the birth of siblings, we are told that, in 1896 at the age of 16, Annie gave birth to Mustafa Mahomet Dost in ‘India’. No further information about this child is provided. In 1898, she marries in Lal Bhaker (the location is not explained and no certificate is available) to Mahomet Dost. Both events happened in modern-day Pakistan. The Dost family had another son and four daughters in Western Australia. Ancestry records her husband’s ‘death’ in 1908 in Port Hedland District, Western Australia (it was actually 1909), and her own 1910 ‘death’ in Karachi, Sindh, Pakistan. As can happen with family trees, she is also recorded incorrectly as marrying in 1939 in Perth!

To support these ‘Life Facts’, documentation is linked from Australia alone. Annie has a birth certificate; Dost does not. Dost has a death certificate; Annie does not. There is no ‘India’ marriage certificate. It is not actually clear if a legal marriage occurred (certainly no ceremony is recorded in Australia, where any Muslim ceremony was unlikely to be legally recognised). It is possible this happened in Pakistan; similarly, their first child, also born in Pakistan, is not documented digitally while those in Australia are.

Other available records reflect how the sexism and racism of the time affect the availability of records. Annie appeared twice on the electoral rolls; as a white, Australia-born woman, she would have been considered eligible. Dost did not appear and was probably considered ineligible to vote because of being an Afghan, despite being from a British colony and being apparently naturalised in 1902 (NAA: A9 1902). This is itself confusing because he was already British. He alone was listed in local city directories twice as head of household, while Annie was never listed. These are exactly the sorts of gendered and racialised silences that one should expect, given record-keeping practices in the early twentieth century. The omissions are only evident in this case, however, because so much other documentation exists.

The banality of what is left is evident in the summary of her life: ‘When Anne Charlotte Grigo was born on 29 February 1880 in Clermont, Queensland, her father, Carl, was 30, and her mother, Maren, was 27. She married Mahomet Dost on 1 January 1898. They had six children in 12 years. She died as a young mother in 1910 in Karachi, Sindh, Pakistan, at the age of 30’ (Ancestry.co.uk n.d.c).

The text is a combination of details from family members and automatically generated text to match the layout that spans every person’s life on Ancestry.com. It is intensely bland, enforcing a high degree of normative storytelling on a highly atypical life. Take the entry for 1908 (which should be 1909), shown in Figure 2 as a screenshot:

This blandness is particularly frustrating in the case of Annie Grigo Dost because there is other information available. It does not need to be like this. Her husband did not simply ‘pass away’. Annie’s story, and that of her family, is clearly complex. It is clearly better referenced than many lives. And yet this context is entirely stripped away. A perusal of any person reveals that the LifeStory function describes death almost always as ‘passed away’ (sometimes the person selected has ‘died’) and is populated mainly with reproducing heteronormative, and imaginarily happy, families. People are ‘born’, get ‘married’, and have babies.

This is unhelpful, not just because it is lazy history, without depth or context, and not just because it makes life boring. Even though Ancestry.com presents itself as the genealogy tool, the guide to who we all really are, the real love and trauma underpinning this family history is erased. In Figure 2, Dost ‘passed away’. According to Ancestry.com, Annie ‘died as a young mother’. The automatically generated language weaves a story that invites us to feel sympathy, but without any depth of historical engagement. Ancestry.com’s sanitised version of the past perpetuates the idea that family, at its core, is something both universal and positive. No family is like this. Annie’s was very different indeed. This combination of silencing and universalising erases both what made Annie special and loved, a figure of historic interest, and the wider power dynamics of her colonial world.

5. Conclusions

Digitisation makes it possible to uncover many aspects of Annie’s life and the stories told about her. But digital platforms perpetuate many of the problems with using colonial records. They also throw in some new challenges too, partly by smoothing out inconvenient parts of family dynamics and presenting them as a ‘LifeStory’ or a ‘Life Fact’.

It is clear that the family discussed here has made efforts to remember this messiness. That deserves recognising. My task here is not to psychoanalyse the family’s traumas per se, but it is impossible to read or hear the evidence of their son, Ameer, and not recognise trauma from witnessing his mother’s death, even if he was too young to remember all the details accurately. This is especially true when he recounts how, years later as an adult, he saw a group of visibly ‘Indian’ men on a street corner, and was immediately filled with terror, convinced they wanted to kill him. As a result, he ran to the nearest police station screaming that they would murder him and had a very embarrassing time when he finally calmed down. Ameer was very aware of why he found Indians frightening, while at the same time missing the Indian food that his sister used to cook and retaining some contact with relatives in colonial India. He clearly took pride in his parents’ beauty and father’s wrestling ability, and (possibly imagined) status as a prince, and yet internalised the racism directed at his family.

When traumatic family histories are impossible to hide, there is an uncomfortable tendency to treat those involved as ‘the inconvenient ancestor’ (Elliott 2017; May 2023; Parham 2008). Ameer and his sisters clearly went through phases of trying to hide their connection to Annie and Dost but reclaimed these connections in later life. It is valuable to see trauma, to ‘bear witness’, as many descendants of Holocaust survivors have argued (Allpress et al. 2010; Elliott 2017; McCabe 2017; Shaw 2021a, 2021b; Robinson-Sweet 2021).

Their personal family dynamics are also revealing of wider historic issues. Cameleers experienced significant racism. Sexism and domestic violence were normalised in Western Australia at this time. The status of men as the primary breadwinners left Annie and her mother as domestics, financially dependent on their husbands, vulnerable to physical abuse and desertion. For Annie, it led to her feeling that she had to go to colonial India to access her husband’s money. Violence, financial hardship, and abandonment were not unique to their family. Their status as a rare ‘mixed race’ family merely exacerbated these wider issues. It made them newsworthy and exotic, but also left the family isolated.

In many ways, this family story embodies the idea of family history as reflecting the wider messiness of national or colonial histories (McGrath 2015; Evans 2022; Allbrook and Scott-Brown 2021). This is a story unique to this particular family and reflective of wider social frameworks, such as hyper-fears in settler societies about white women’s sexual relationships with ‘others’, and the ambiguous position of offspring racially. The context of all of this is important for this family’s history and for understanding race and colonialism and gender at the time. Instead, a bland family narrative is imposed on platforms like Ancestry.com. If colonial archives were part of ‘an imperial project of domination and affirmation’ (Mueller 2017), with Ancestry.com we can see a different form of domination, one that silences family experiences that are not ‘feel good’.

We do not know exactly how Ancestry.com’s algorithms work, how such stories are generated, nor do we know the degree to which this shapes people’s understandings of the past. We do not know how they select which records to highlight and which to not highlight. It seems unlikely that an employee actively decided to hide those particular murders. Instead, Robinson-Sweet’s and my examples suggest that Ancestry.com has an algorithm that glosses over anything that might generate awkwardness in the user, adopting a particularly bland language to describe things like death and making assumptions about which documents to link and share. Such a default is potentially highly damaging to our ability to understand the past. As family historians, whether professional or not, this is a lesson in how it can be necessary to ‘purposefully disrupt comforting and secure ways of remembering and knowing’ (Byrnes and Coleborne 2023). Otherwise, glossing over this history makes it impossible for families to recognise and address trauma caused by violence, domestic abuse, racism, and colonialism, or for historians to understand the lived experiences of these families. If family histories tell us who we really are, then such erasure is actively damaging.

Adult researchers, whether professional or not, may think that they will know how to ‘read’ Ancestry.com and similar platforms and know not to take it as definitive. Sometimes, yes. But as I have outlined here, we do not always know what is missing or how what we see is being filtered. Also, not all users are adults. In 2022, Ancestry launched AncestryClassroom™ for teachers and educators in Australia and New Zealand. This is how they advertised their new service:

For nearly a decade, AncestryClassroom™ has helped students learn more about themselves, navigate the world around them, and become more resilient. We do this by providing educators with access to classroom resources, professional learning tools and Ancestry historical record collections at no cost.

In an effort to reach more students globally, Ancestry is thrilled to expand access to AncestryClassroom™ to educators and students in Australia and New Zealand.

More than 7 million students across the United States currently have access to AncestryClassroom™ and we are excited to support educators in Australia and New Zealand in helping their students personally connect to history, develop research skills, access original documents, develop a sense of self and much more.

AncestryClassroom™ Australia and New Zealand includes access to Ancestry content from a UK Heritage Plus membership, for free. This includes access to all Australia, New Zealand, UK and Ireland records.

(Ancestry.com 2022)

Here is a promise that accessing this history will help students develop their own sense of ‘self’, of identity. They will also gain valuable skills navigating documents and the ‘world around them’. And yet, this can be only a partial image of the world and even of family histories. It also draws upon the basic assumption that Australians and New Zealanders are ‘from’ Britain and Ireland. Annie and Mohamot Dost would be missing from this collection, as neither had ancestors from Britain or Ireland. One wonders what other omissions or decontextualisations occur.

I do want to emphasise that Ancestry.com is not the only company to suffer from problematic digital platforms and presentations. However, they deliberately market themselves as the largest and best-resourced service. They stress how the data can tell the users more about themselves through uncovering their family histories. And, relatively uniquely, their platform imposes a story onto the data, a story that is presented as authoritative while being also quite superficial. Most platforms do not market themselves in quite such a totalizing way. Instead of that totality, we are given an anodyne, simplified presentation of a past stripped of anything uncomfortable or interesting. It importantly takes away the chance to confront or acknowledge complex histories. We are sold an image of reclaiming lives like Annie’s from the past, but what is actually offered is deeply problematic.

This article calls on both Ancestry.com and its users to reconsider assumptions about why users engage with family history, and what stories we may want to hear. There is already a range of scholarship that has explored how and why families may seek out traumatic family histories. For instance, they may ‘seek self-understanding throughout establishing a meaningful personal identity and place in the world’ through family history research. It may even relate to the ‘secular pursuit of belonging to something larger than the individual self’. It can also cause distress and ambivalence (Krauskopf et al. 2023). Yet there is a power in ‘bearing witness’, as any social historian will tell you. Digital platforms should empower family historians; Ancestry.com and FindMyPast claim to want to do this. Now we need digital platforms to actually do this.

We also need to consider how private companies and public platforms can signpost related connections more clearly. Even for a case as well documented as this, tracking down digitised sources took considerable effort. No doubt, many non-digitised sources would also be useful (especially in India or Pakistan or Indian colonial records).

There also needs to be more openness around all digital archiving platforms about the archival silences that can happen through the process of the digitisation and organisation of data. This is not a neutral process, but one which can too easily erase or gloss over the real complexities of the past, especially when data are re-organised by individual, by hetero-normative family structures, and names are organised in a European style. The past is messy. We are messy. If we want to use the past to understand ourselves and our worlds, we deserve a more open recognition of the potentials and pitfalls of digital methods.