Official Languages Committee on April 13th, 2016

[The Chair, Liberal, Denis Paradis]

Welcome to this meeting of the Standing Committee on Official Languages. Pursuant to Standing Order 108(2), we are continuing our study on the Translation Bureau.

We have with us Linda Cardinal, who is a titular professor with the School of Political Studies at the University of Ottawa, and Jean Delisle, professor emeritus of the University of Ottawa, who is appearing as an individual.

I would like to welcome you both.

In the first hour, you will both have about 10 minutes to make your presentations, and then we will have a period for questions and comments from committee members.

Please go ahead.

[Linda Cardinal, Titular Professor, School of Political Studies, University of Ottawa]

Thank you very much, Mr. Chair and members of the committee.

Thank you for inviting me to offer my thoughts on the controversy surrounding the implementation of the Portage tool in the federal public service. I am delighted to be here today, as the debate about the Portage tool enables me to raise issues that go well beyond the deployment of the tool, but that the tool has unexpectedly crystallized.

Let me say at the outset that there is a real threat at our door, and that door leads to the federal public service, and to the Translation Bureau in particular. Where there is a threat, there is also an obligation to take swift action.

The debate on the implementation of the Portage tool has revealed that official languages are suffering in the public service, particularly the standing of French.

I will argue that we need to review the role of language technologies and better understand their impact on linguistic duality. We also need to rethink the Translation Bureau in order to give it the means to achieve its objectives and reverse the trend toward de-skilling translation professionals. And the decision to deploy the Portage tool must be rescinded.

First, I want to tell you about my experience as a former member of the working group on government transformations and the official languages, which was established by the Treasury Board in 1998-99, as that experience will provide the context for my remarks.

In 1998, the federal government launched transformation efforts to resolve its budget deficit problems. The government reviewed its programs and methods of delivering public services, resulting in a fundamental re-evaluation of the scope of government intervention in society.

The government transformations of the time included the creation of the Canadian Food Inspection Agency, the Parks Canada Agency, the Canada Business Centre and the Canadian Tourism Commission. In addition, corporations such as Petro-Canada, Air Canada, the country's airports, Canadian National Railway and Nav Canada were privatized.

The increased use of information technology in organizing public services was expected to improve the delivery of services to the public. Adopting new technologies would lay the foundation for a more effective government that was better connected with its citizens yet benefited from economies of scale.

In 1998, after condemning the government transformations, the Commissioner of Official Languages at the time, Victor Goldbloom, called for a working group to be established. The federal government accepted the commissioner's recommendation and ordered the Treasury Board to create the working group—on which I served—to study the impact of the government transformations on linguistic duality, particularly as regards Part VII of the Official Languages Act.

In carrying out our mandate, we found people were tired of dealing with official-language minority issues, we confirmed the concerns raised by the commissioner, we saw in very concrete terms the negative impact of the government transformations on official-language minorities and we proposed ways to turn things around. The working group made a number of recommendations in its report. For example, it asked the Treasury Board to report on the impact of the government transformations on official languages, create a mechanism for consulting with official-language minorities as part of the review of the transformations and continually remind the institutions subject to the Official Languages Act of their obligations.

Unfortunately, the working group did not raise the issue of the Translation Bureau at the time. I see today that the bureau's difficulties also began around that time. These problems grew under the previous government, which did not hesitate to cut staff and privatize services that support official languages in the public service, such as the privatization of language courses, which can be assumed to mean French courses.

Following the working group's review, the Action Plan for Official Languages was published in 2003. It included measures to promote research into language technologies. At the time, the action plan was a new tool to help the government meet its obligations to support the development and enhance the vitality of Canada's official-language minorities. However, today we have an impoverished vision of official languages. This vision is based on a utilitarian logic that puts private interests ahead of the public interest. Here are some examples.

In Blueprint 2020, the government explains that it must be innovative in the realm of technology and equip public servants with technological tools. Yet, in 2016, the media have revealed that the Translation Bureau has abolished 400 positions since 2010. Is that accurate? Moreover, 140 more positions will be eliminated by 2017–2018. It is fair to ask whether equipping public servants is really about promoting the use of both official languages or whether the government is gradually shuttering the bureau.

Let us turn to the Portage tool. I fear that the implementation of this tool not only violates the Official Languages Act, but also reveals an ignorance of the issues associated with translation. On the one hand, the government thinks translation is important because it wants to provide a tool that everyone can use. On the other hand, the government is devaluing the role of translators and the specialized knowledge they possess. The government seems to believe that everyone can translate even though there are inherent requirements for translation that only translators have mastered. If I were to venture a parallel, it would be to journalism and social media. On Twitter or Facebook, anyone can call themselves a journalist. But we know full well that the journalism profession is demanding and requires hard work, thoughtfulness and writing skills. Twitter's 140 characters will never replace investigative journalism, analysis or lengthy editorials.

The Portage tool is a troubling sign for official languages. If public servants need such a tool, it is because they do not know enough French to write emails or draft internal communications. What happened to promoting linguistic duality within the federal public service? I ask this question because of the asymmetrical position of two official languages in the public service. After 10 years of Conservative government, what is the status of French in the public service? What has happened to Part V of the act and public servants' right to work in the official language of their choice?

I propose two recommendations in order to formalize my arguments and encourage you to take action.

My first recommendation is that the Government of Canada reverse its decision to implement the Portage tool.

Granted, this is not an original recommendation, but it is a necessary one. I join many other translation and official-languages stakeholders in calling for a reversal. The new government is not bound to the decision of the previous government, especially when it comes to linguistic duality.

My second recommendation is that the Government of Canada establish a working group on the status of official languages in the public service and that this working group devote particular attention to the role of language technologies in promoting linguistic duality, the situation at the Translation Bureau and the impact of the privatization of services, such as French courses, on official languages.

My recommendation is ambitious, but essential. A working group on the status of official languages in the public service must reaffirm the language rights of francophone public servants and the right of francophones to receive communications in French that has not been produced by a machine, as my colleague Jean Delisle emphasized recently in Le Droit.

I agree with Jean Delisle that respecting language rights means ensuring idiomatic French and that we must not let French be turned into a robot language. In an asymmetrical context like the public service, where French is a translated language—85% of French documents are translations—we cannot sit on our hands while the government deploys a tool that could reduce French to a bastard language.

One way to change the status of French in the public service would be to encourage francophone public servants to prepare documents in French in order to counter the asymmetry and recognize their right to draft documents, messages, notes and presentations in French.

Thank you for your attention. I will answer your questions later.

[The Chair, Liberal, Denis Paradis]

Thank you very much, Mrs. Cardinal.

We now go to Mr. Delisle.

[Jean Delisle, Professor Emeritus, As an Individual]

Thank you, Mr. Chair.

Thank you for inviting me to appear before your committee to participate in the discussion about the Translation Bureau. I hope to provide a relevant, if not original point of view.

I myself was a translator at the Secretary of State early in my career, and I recently finished a book on the Translation Bureau's first 100 years entitled Les douaniers des langues: grandeur et misère de la traduction à Ottawa, 1867-1967.

In 1984, I published a history of the bureau on the occasion of its 50th anniversary. I have dedicated some 40 years to translation history and teaching, which supports my testimony.

We all know that a lot of translation goes on in Canada. Translation is part of this country's DNA, even though many Canadians consider it to be a necessary evil of Confederation. The same could be said of official bilingualism because translation and bilingualism go hand in hand. Translation is not a by-product of bilingualism; it is a manifestation of bilingualism.

What should we think of the Portage machine translation software? Modern technology is marvellous, but it must be used wisely. We need to differentiate between the availability of Portage to all federal public servants and its use by professional translators.

Let's start with federal public servants.

There are risks associated with machine translation software such as Google Translate and Portage in their current incarnation because they are unpredictable and unreliable and there are no clear guidelines for their use by federal public servants. Canada's three largest translators' associations—Ontario, Quebec and New Brunswick—have expressed significant concerns about Portage. Even the government's informatics services have not recommended using it for email communications. If Portage were to become widely used, it would signal a return to the days when translation was entrusted to bilingual secretaries who were not equipped to perform such tasks.

We know that, for several years now, public servants have been using Google Translate for internal communications, often with disastrous results. I provide some real examples of these kinds of emails in my brief.

Canada's official languages are English and French. Machine language is not an official language, and neither is it “good enough”. For whom is that language supposed to be good enough?

Because both languages are equal under law, Portage could produce errors with legal consequences. Such errors could discredit the Translation Bureau and tarnish Canada's image.

I think it's unrealistic to believe that machine translation can improve communication among public servants or encourage them to work in their language. What is preventing them from working in their language right now?

A unilingual public servant who is unable to read the original message could be misled by an unrevised machine translation. If all Portage translations must be revised, there will have to be a considerable number of bilingual revisers, and that entails extra cost. Will the government really save money? Who will those revisers be? What skills will they be required to have? These are some questions that will need to be answered.

What about professional translators? For over a century, federal translators have specialized in language and translation. Because of them, Ottawa has become a centre of excellence for translation and interpretation. They professionalized the profession in the 1920s and pioneered translation teaching in the 1930s.

Lawmakers in three provinces have granted the reserved title of certified translator to translators who are members of professional associations in those provinces. They are recognized as professionals. Translators play an essential role in ensuring Canada's linguistic duality and practise a profession with a strong symbolic value. They are considered by many MPs, ministers and senators as the cement of national unity. Dozens and dozens of testimonies attest to that.

Their primary complaint is the lack of respect for their professional status: they are being told not only what to do, but also how to do it. Decisions about how to work must remain the professional's prerogative. The fact that they are public servants is irrelevant. Does anyone tell accountants how to perform their work?

Federal translators are certainly not hostile to new technology, but they are very aware that these new tools, Portage in particular, could have a detrimental effect on language, translation and translation as a profession. They are willing to use machine translation, but they do not want to be turned into machines themselves.

Let's look now at the Translation Bureau, which seems to be at a crossroads.

The Translation Bureau Act, enacted in 1934, requires the organization to make and revise all translations for all government organizations. In the years following the Official Languages Act, the bureau flourished. The bureau expanded the range of services available to public servants and all Canadians. In 1974, it even acquired oversight over linguistic standardization within the federal government.

As a public organization, the Translation Bureau has demonstrated dynamism, leadership and innovation in the last 50 years. My brief contains an impressive list of these innovations.

The Translation Bureau is responsible for producing quality translations, but the law does not say that this must be done at the lowest possible cost. Quality comes at a price, but translation is actually not very expensive.

Historically, the bureau's budget has always represented less than 1% of the national budget, which is currently $296 billion. Is accounting expensive? How about consultants? We rarely ask the question in those terms, but when it comes to discussing translation, the financial aspect is raised immediately.

Over the past 10 years or so, there has been a clear desire to reduce translation-related expenses as much as possible. There is talk of cutting bureau staff by 60%. Over the past three years alone, the bureau looked to save $50 million thanks in some measure to new technology. Is it purely by chance that this cost-cutting coincided with the plan to roll out machine translation software on April 1?

We need to give control over all translation within the federal public service back to the Translation Bureau in accordance with the act that has governed it since 1934. If not, the anarchic system that prevailed in the pre-bureau days will return. I think we're already there. I could answer that question if you ask it later. I could tell you how we currently have the anarchic system we wanted to avoid in 1934.

The bureau is a public organization and, as such, it has a mission that differs from that of a private translation company. I could also explain that more later.

I may be mistaken, but for a number of reasons, I get the strong impression that there are forces seeking to dismantle the Translation Bureau. The bureau is not recruiting; its workforce has been shrinking steadily. It has offered no internships in four years and has stopped funding the Traduca internship program. Is this because it foresees a need for fewer translators?

In addition to excellent interpreters who accomplish masterful feats of communication on a daily basis, the bureau had a team of terminologists who literally invented this new profession and whose remarkable achievements have garnered worldwide recognition. However, I'm told that the number of terminologists has shrunk to the point that terminology is practically endangered and significant expertise has been lost, as significant as the loss of scientific and technical translators who must retire or who are being fired.

I believe that the bureau's status as a special operating agency—or SOA—prevents it from fulfilling a public organization's mission in terms of innovation, training and terminology. As in any field, failure to innovate means falling behind. The Translation Bureau's history is part and parcel of the evolution of our bilingual nation. It reminds me of a grand heritage building that developers want to demolish for the sake of financial gain.

To conclude, I'd say that translation is an excellent indicator of the relative standing, weight and vitality of one official language vis-à-vis the other. The first language to suffer the detrimental effects of machine translation would be French, which is the main language that is translated. There were 325 million words translated into French compared to 23 million into English in the last fiscal year. Francophones in this country will reject cost savings and productivity as excuses to relegate their language to the ranks of technobabble.

Thank you. I would be pleased to answer all your questions.

[The Chair, Liberal, Denis Paradis]

Thank you very much for your presentation, Mr. Delisle.

Thank you, Mrs. Cardinal.

Since we are running a little behind, we will go directly to the first round of questions, with six minutes for each member.

We'll start with John Nater.

[John Nater, Conservative, Perth—Wellington, ON]

Thank you very much, Mr. Chair.

I would like to thank the witnesses for their presentations.

As you can hear, French is not my mother tongue. I am taking French courses, but I don't have the vocabulary to speak fluently in French. So Google Translate is my best friend. I use it often. It's an aid for me, and for my colleagues. I was a public servant at the Treasury Board for two years, and I would have liked a translation tool.

For me, any tool that can help promote bilingualism is useful, especially for those of us who may not have had the advantages of French immersion or the opportunity to use our second national language when we were growing up and had to find alternative means to use our second language.

We've heard opposition to tools such as this from a couple of witnesses now, and I'm a little concerned about that. It will never replace translators or the professionalism of translation. I accept that 100%. We've used the services of the Translation Bureau in our office. I used them as a public servant, and they provide exceptional quality. There's nothing I could say wrong about that.

I want to focus on a couple of things that Madame Cardinal spoke about.

The comparison to social media is actually a worthy comparison, but I come to a conclusion that is different from yours. Yes, there are differences between social media and journalism, but they go hand in hand. We see the tools and the way in which social media are being used as a journalistic tool in the same way that professional public service-endorsed translation tools could find worthwhile use in the public service.

Do you see any role for a machine translator tool that could be used by public servants themselves in our day-to-day operations or in our day-to-day lives as public servants?

[Linda Cardinal]

Thank you for your question.

I didn't say that all tools were bad. We use many tools in our life, including to wash clothes, clean the house or read. You have a tablet. Technology is a part of our life. No one here is saying that we need to go back to the stone age; even then, people developed tools. Tools are part of us, and we cannot do without them, but it depends on the context in which the tool is used because a tool is not neutral. Right now, the approach for the proposed tool is a utilitarian one that aims to gradually get rid of translators, to replace the translators.

Even if social media makes people want to become journalists, it isn't in the process of replacing the profession of journalist. Just because I use a tool to do my accounting, that doesn't mean there won't be any accountants tomorrow. Just because I am versed in law, that doesn't make me a lawyer. We can make these parallels with professions ad infinitum, and we will always come back to the same thing, which is that the tools aren't there to replace people, the professionals; they are there to help do the work.

Certainly, in the past, some tools have meant that businesses needed fewer workers, among other things.

I'm not here to give you a background on technology, but in this case, it's also important to look at the context and the purpose behind these tools. Then we have to see how a policy in this area interacts with the Official Languages Act. I have the impression in this case that the interaction is jammed. In other words, some aspects don't go together. People want to use a tool for utilitarian purposes to promote official languages, when the goal of official languages is to enhance the vitality and development of official languages and to encourage their use. That isn't the case here.

[John Nater, Conservative, Perth—Wellington, ON]

I accept your argument, but—actually, I don't accept the argument. I don't think anyone is arguing that professional translators will become irrelevant. Quite the contrary; there's an essential role that professional translators must play in translation.

A machine tool will never be used to translate any number of official documents or legalistic documents, but it is an essential tool, I would say, especially—and again, I give credit to Minister Brison for his promotion of more millennials in the public service.

A machine tool, an online tool, a computer program for translation, I would argue, is a promotion of the French language, or of both official languages. I'm not saying there are not challenges. I know when I use Google Translate, I have trouble with the translation, but what I am saying is that it is essential to have that option of short translation with a tool. We don't bemoan the loss of telegraph machines, and in the same way we evolve. There is an essential use for an online tool such as this.

I want to touch on something else, though. You brought up the idea that a tool such as this may violate the Official Languages Act. I was hoping you could expand on that. How exactly would the use of an online translator, an online tool or computer tool such as this, violate the Official Languages Act, especially when it's not necessarily being used to translate official documents?

[The Chair, Liberal, Denis Paradis]

Please give a very short answer, Mrs. Cardinal, because we have gone over time.

[Linda Cardinal]

I will answer very quickly then.

I think translators need to be given tools to help them do their job. What you are saying runs counter to the solid evidence that shows that the Translation Bureau's workforce is continuously shrinking.

Internal work isn't innocuous. Short emails aren't innocuous. You're all on your iPhones, and you can send thirty, forty, fifty emails a day. Writing all those emails in French, or at least even some of them in French, is still a lot of work.

For the last point, I think there's a problem in terms of the Official Languages Act. Under the Official Languages Act, French and English must be treated equally. A machine language and an idiomatic language are not equal. There is an asymmetry of the French and English, and that's the problem French faces in the federal public service. You will increase this asymmetry; you won't be helping to reduce it.

[The Chair, Liberal, Denis Paradis]

Thank you, Mrs. Cardinal.

Let's move on to the next person.

Mr. Lefebvre, I believe you are going to share your time.

[Paul Lefebvre, Liberal, Sudbury, ON]

Yes, I'll share it with Mr. Fergus.

I'd like to thank the witnesses for being here today.

You make a few strong conclusions with the title of your brief: “Linguistic duality imperiled by utilitarian logic in the federal public service.”

Mr. Delisle, you said, “Francophones in this country will reject cost savings and productivity as excuses to relegate their language to the ranks of technobabble.” Those are strong words.

You also said that, in your opinion, the purpose of this technology is to replace humans.

What are you basing this statement on? Is it a fact, or are you assuming that? Where does that comment come from?

[Linda Cardinal]

Who is the question for?

[Paul Lefebvre, Liberal, Sudbury, ON]

For you, Mrs. Cardinal.

[Linda Cardinal]

Okay.

I said this during an exchange with Paul Gaboury in Le Droit. When we look at the decline in the Translation Bureau's staffing levels over several years, we have to ask the question: is it to replace humans? There are humans behind the tools. Are the tools being used only to replace translators?

Once again, I come back to the idea that everyone thinks translation is important. That's why people want tools. At the same time, why is a translation tool used to replace translators, while an accounting tool isn't used to replace accountants?

[Paul Lefebvre, Liberal, Sudbury, ON]

How do you think this translation tool should be used? What is the ultimate goal of this tool?

[Linda Cardinal]

A translation tool is used to help the translator. The tools we use should not have an official status to replace translation work and change translators into post-editors.

The translation profession needs to be protected. Mr. Delisle can also tell you about that. Canada's image is at stake. It's important to remember that language is not economics; language is politics, language is used to unite this country. If we start fiddling with official language issues for utilitarian reasons, the country's unity will be at stake. This isn't because of Quebec; it's because French and English are Canada's two official languages. I don't have to convince you.

[Jean Delisle]

I'd like to add something if I may.

It all depends on how the tool is used. One person uses it to understand certain texts. That isn't a problem. It becomes a problem the moment someone uses the tool to communicate messages or to do published professional translations.

You have here the employee separation document for a public service employee. The machine was used to translate one part into French. It reads “L'achèvement de la “émis” des composantes de ce formulaire”. Do you understand anything? The form is printed in English on one side and French on the other.

This is what we want to condemn, using the machine to publish texts.

[Paul Lefebvre, Liberal, Sudbury, ON]

That was why I asked you what you thought the purpose of the tool was, and if the tool was used for external communications.

I think Mr. Fergus will continue now.

[The Chair, Liberal, Denis Paradis]

First, though, I'd like to ask you, Mr. Delisle, to provide the clerk with the document you just referred to.

Mr. Fergus, you have the floor.

[Greg Fergus, Liberal, Hull—Aylmer, QC]

I'd like to thank the witnesses for being here.

I fully agree with your basic analysis and what you both propose in your presentations. Perhaps it's because I grew up in the western suburbs of Montreal. I was the only Black person in a White neighbourhood. I was the only anglophone among many francophones. So I'm quite sensitive to the issue of minorities.

What I like about your contribution to this discussion is that you are asking what the purpose of these tools is, what the purpose of translation is. I think the purpose is to ensure that anglophone or francophone public servants have the right to express themselves and be understood in their own language. So I don't fully agree with your recommendations, but I do with most of them.

Mr. Delisle, you mentioned that modern technology has accomplished marvellous things, but it must be used wisely. I think you're absolutely right. What kind of guidelines do we need? I maintain that it is important to use this tool as a tool for understanding and not for translation. We know quite well that if people start doing that, francophones will start writing messages in English.

[Jean Delisle]

I fully agree with you.

One of my recommendations is that the roll-out of the software be suspended until further notice so a study can be done on the consequences it may have and how public servants might use it. If it is to read a message and understand a text, that's not a problem, but if it is to publish things, that becomes problematic. Guidelines should be established to have the texts revised by qualified people.

[Greg Fergus, Liberal, Hull—Aylmer, QC]

It's important to mention that this applies to both official communications and informal communications.

[Jean Delisle]

Absolutely.

In some way, a guide on the proper use of technology should be developed. That's what I think about it.

[The Chair, Liberal, Denis Paradis]

Thank you very much.

Mr. Choquette, you have six minutes.