Friday, September 26, 2008

A Lesson in Writing from 1944

I watched the Presidential debate tonight. One of the candidates mentioned a pair of letters that General Dwight David Eisenhower wrote in 1944. He wrote one letter that he would use in the event of a victorious Normandy invasion, and he wrote another one that he would use in the event of a defeat.

I was curious about those letters, so I googled for them. I found something interesting in a way that I didn't expected. Here's the text of the letter that General Eisenhower wrote in case the invasion force at Normandy had been defeated:
Our landings in the Cherbourg-Havre area have failed to gain a satisfactory foothold and I have withdrawn the troops. My decision to attack at this time and place was based on the best information available. The troops, the air and the Navy did all that Bravery and devotion to duty could do. If any blame or fault attaches to the attempt it is mine alone. —July 5
Here is a picture of the handwritten note, which I found at

The handwritten note contains some important information that isn't present in the transcribed text alone. Observe that General Eisenhower edited his message. He actually edited himself three times; I'll refer here only to the top one. Here's the original version:
Our landings in the Cherbourg-Havre area have failed to gain a satisfactory foothold and the troops have been withdrawn.
Here's the modified version:
Our landings in the Cherbourg-Havre area have failed to gain a satisfactory foothold and I have withdrawn the troops.
The difference is subtle but important. In grammatical terms, General Eisenhower made the choice to discard passive voice and adopt the direct, subject-verb-object style of active voice. One Wikipedia article that I particularly admire identifies passive voice as a tactic of weasel wording: "Weasel words are usually expressed with deliberate imprecision with the intention to mislead the listeners or readers into believing statements for which sources are not readily available."

In Eisenhower's original version, he had stated that "the troops have been withdrawn." From this statement, we would have learned some information about the troops, but we would not have learned directly about who had withdrawn them. This passive-voice language, "the troops have been withdrawn," would have subtly conveyed the notion that the author wished to conceal the identity of the decision-maker about the withdrawal.

In the modified version, General Eisenhower made it abundantly clear who had made the decision: he did. The revised wording is more informative, it is more efficient, and it is more courageous.

Active-voice writing holds several advantages over passive-voice writing. I've learned this in my work, especially in consulting engagement reports, where I've found it's essential to write with active voice. Advantages of active-voice writing include:
  • Active voice transmits more information to the reader.
  • Active voice is plainer and simpler; it is easier to read.
  • Active voice is often more economical; it conveys as much or more information in fewer words.
  • Active voice is often more courageous.
The value of courage is obvious in the Eisenhower case. Even if the Allies had been defeated at Normandy, Eisenhower was courageous enough to accept the responsibility for the plan, its execution, and even its remediation.

Courage is also important in our writing about technology. Writing with active voice can be much more difficult than writing with passive voice. ...Because, you see, active voice gives you noplace to hide. When you know something, you say it. When you don't, active voice writing pretty much forces you to say that. It can be quite unsettling to admit to your audience that you don't know everything you wish you knew. It takes courage.

If you find yourself ashamed that your writing is too vague or that it asks more questions than it answers, then I think you have only four choices. (1) You can decide not to write anymore because it's too hard; (2) You can try to conceal your deficiencies with weasel wording; (3) You can admit the gaps in your work; or (4) You can improve the quality of your own knowledge.

Of course, I don't believe that giving up is the right answer. Option two—concealing your deficiencies with weasel wording—is, I think, by far the worst option of the four. Choice three frightens a lot of people, but actually it's not so bad. I believe that one of the great successes of the modern wiki- and forum-enabled Internet is the ease with which an author can voice unfinished ideas without feeling out of place. The fourth option is a fantastic solution if you have the time, the inclination, and the talent for it.

Back to General Eisenhower's note... I find his edit inspiring. By making it, he reveals something about his thought process. He wrote his original text in the common, politically safe "tasks have been executed" kind of way. But his edit reveals that it was especially important to him to be direct and forthcoming about who was making the decisions here, and who was at fault in case those decisions went wrong.

Knowing that General Eisenhower edited his note in the particular way that he did actually makes me respect him even more than if he had written it in active voice in the first place.

* * *

Here's where I thought I was finished for the evening. But I want to show you what it looks like to execute faithfully upon my own bitter advice. Eisenhower's letter piqued my interest in the D-Day invasion of Normandy. One thing I noticed is that the invasion was initiated on June 6, 1944. Eisenhower's memo is dated "July 5." Uh, that's a month after the invasion, not the night before. It was another hour or so of writing lots more stuff (which I've long since deleted) before I googled "eisenhower message june july" and found this, which states simply that, "The handwritten message by General Eisenhower, the In Case of Failure message, is mistakenly dated 'July' 5 instead of 'June' 5."

Ok. I can accept this as authoritative for my own purposes, for one, because it doesn't matter too much to me tonight if it's not true. It's a plausible mistake to imagine a man making who's under as much pressure as he would have been on June 5, 1944. For comparison, I could barely remember my own phone number on the night of the Loma Priete earthquake, which I rode out in the Foster City Holiday Inn in 1989. But of course, such an anecdote about me is no proof of this particular proposition about Dwight D. Eisenhower.

So, do you see what I mean when I say that writing is HARD!? The act of writing itself—if you try to do it well—forces you to do work that you never intended to do when you set out to write your piece.

That's one of the good things about the software industry. When someone makes a statement about computer software, I can confirm or refute the statement myself using strace, DTrace, 10046, block dumps, or some other research tool that I can actually get my hands on. That doesn't make it easy, but it usually does make it at least possible.


Pete Scott said...

In defence of the passive.
I am often too passive in my writing, but I blame that on my education - being a chemistry graduate and one-time research scientist I was trained to write about the events and the observations and not about the observer. However, observation and reporting of fact
(but not necessarily postulating reasons) are passive acts and not at all the same as responsibility and leadership that should be exhibited by politicians where the non-passive does, as you say, indicate courage

Anonymous said...


"then I think you have only four choices. (1) You can decide not to write anymore because it's too hard; (2) You can try to conceal your deficiencies with weasel wording; (3) You can admit the gaps in your work; or (4) You can improve the quality of your own knowledge."

Is it not required to admit the gaps in your work, before you can improve your knowledge? How else would you know what you must improve?

Cary Millsap said...


Even in the sciences, though, we can use passive voice to cover our own gaps. I don't have a good chemistry example off the top of my head, but I do remember an example from my early days at Oracle (early 1990s).

I can remember not knowing explicitly which process it was that did the work of row retrieval. With passive voice, it was possible to discuss row retrieval for years without actually knowing which process it was that was executing the work. (The rows had simply "been fetched," etc.)

It took some Unix work (ps, strace, etc.) before I could really construct confident SVO (subject-verb-object) sentences about which process was doing what. With tools like pstack (see Tanel's work for example), you can know even which specific function calls within the server process are doing which units of work.

You just can't write SVO sentences without knowing what S is. Passive voice is a tool that people can use (well, abuse) to completely conceal the relationship between S and O without a lot of people even noticing. Often, that relationship is more informative than anything else about the sentence.

Cary Millsap said...


Oh, you're right square on target. Definitely. But in my experience, many people don't want to learn that there's anything about them that needs improving (myself included some days). That's where courage comes in.

Learning is a state transition between not knowing something and knowing it. For someone to truly learn, he has to have the courage to admit—at least to himself—that there's something he doesn't know. Some people are unwilling to do that.

Good writing can act as a forcing function to make learning happen for the author. That's an important distinction. A lot of people view writing as "merely" the act of recording what you already know. I view writing—good writing, anyway—as much, much more than that. ...For exactly the reason you have identified.

Yet Another Mother Runner said...

1. The Palin-Couric interview wreaked of passive voice on Palin's part last week :)

2. Where did you start looking to turn your passive voices into active voices (related to your early years in Oracle)??

Cary Millsap said...


I haven't looked recently at a lot of my early writing. I hope there's not an over-abundance of cringe-worthy material in there. :-)

To answer your question, I think I really started paying attention to the active/passive issue in the later 1990s when I was reviewing a lot of my staff's consulting engagement summaries.

It's one of those issues that I had known about for a long time, but I didn't begin to really appreciate the difference it could make until I had to confront it professionally every day.

Yet Another Mother Runner said...
This comment has been removed by the author.
Yet Another Mother Runner said...

Thanks, Cary!
If you happened to stop by my blog, it talks about all but my professional life...(not yet at least :)
I am an Oracle (pl/sql, apex) Developer among other things...

Anyways, in your opinion -
What Should I do to get there?
How do I know what I do not know yet?
I know I can find answers when I face an issue. But how do I find out if it is the right way?

Cary Millsap said...


What Should I do to get there? Step 1: Find interesting problems.

How do I know what I do not know yet? The easiest way is to put yourself in a situation where people are counting on you to find out answers to interesting problems. They will load you up with interesting questions. Over time, you may begin to generate more interesting questions yourself than other people will.

I know I can find answers when I face an issue. But how do I find out if it is the right way? Learn how to measure the Right Things. I believe that the Right Things are things that are close to the customer. In software performance, for example, I believe that response time is more of a Right Thing to measure than utilization.

Yet Another Mother Runner said...

Thanks again!
Be sure to look for questions from me when I'm trying to look for answers ;)

Jared said...

This strikes a responsive chord in me.

For years in school (starting in grade school) it seems I was encouraged to write in passive voice.

"Too many I's" was a bad thing, the point being that I was trying to draw too much attention to myself as the writer, rather than the observer.

Maybe it was over-impressed on an impressionable young student, as
I still have a problem with that, sometimes going back over my writing (blog, email, whatever) and editing from active to passive.

I think I'm done doing that. Thanks for this blog post Cary, it was quite insightful.

Anonymous said...

Hi Cary

Great post !!

It rings so very true from my perspective. I thought I might add some of my thoughts and opinions on my blog (in as active voice as I could ;)

Well done !!


Paul Moore said...

It's interesting that the note on the error in the date was itself in passive voice - "the message ... is mistakenly dated". Having just read your post, I immediately thought "who made the mistake?" and I was less prepared to accept the authority of the comment as a result.

Cary Millsap said...


Great catch...

Here, the author could have more directly said, "General Eisenhower mistakenly dated his note 'July' 5 instead of 'June' 5." That tells a clearer story with fewer words, doesn't it? And it actually does a better job of asking an obvious next question like, "How could General Eisenhower have made such a mistake?" ...Which would be an interesting fact to add, in my opinion.

I think many authors choose the passive voice in situations like this because it sounds less indicting, less confrontational. ...More polite. But I think sometimes, people choose to write with passive voice specifically because it creates less pressure to expand the scope of an article to things the author may not know or want to write about.

Mohamed Houri said...


Excellent!! I have been reading this blog article for several days and each time I read it I feel that I haven't read it before. It is very sad for me that I can't find such kind of excellent articles in my mother tongues (Arabic and French) so that I can make my comments and my "internal" feeling in perfect synchronisation.

I don't think that employing active writing requires courage for those who will always consider them selves in a continuous learning process. If I made a mistake when employing active writing I will either be in one of the two situations (1) no one will point out this mistake or (2) my mistake will be pointed out and I will consider that I learned something new which will make my mistake not reproducible in my future active writing.

May be you have remarked that I didn’t employed a passive writing here and will from now on keep in this style of writing

Thanks a lot!


PS : Normandie (Normandy)

Anonymous said...


What an inspiring post.

Seeing a letter that shows one man stating categorically "I am responsible" for such an enormous eventuality is quite spine-tingling; and, as you say, seeing the deliberate change from "have been withdrawn" to "I have withdrawn" rams the point home even more clearly.

I also thought your commentary on active voice vs. passive voice, and the role of the passive as a tool for deception was masterly.

However, I would like to stand up for the passive voice just in case any of your readers get too keen on condemning it out of hand.

The general structure of the passive is: "subject verb agent". In my view, the use of the passive for "weaseling" depends largely on omitting the agent. It's not the appearance of the passive itself.

Here's a weasel use of the active:
"... and the troops have withdrawn."

Here's a non-weasel use of the passive - which makes the responsibility very clear:
"... and the troops have been withdrawn on my orders".

Here's another non-weasel use of the passive that (for subjective reasons) makes me WANT to use the active because the passive FEELS so clunky:

"... and the troops have been withdrawn by me."

The passive is harder to use well, it can be harder to read, and it does make it easier to avoid responsibility, confrontation, or understanding the topic - but it is not inherently a bad thing.

Or, putting it the other way round: people often use the passive badly, people often find it harder to read the passive, and responsibility, confrontation and the requirement to know your topic are easier to avoid if you use the passive.

Jonathan Lewis

Brian Tkatch said...

The Elements of style explains Active vs. Passive voice.

Cary Millsap said...

Jared, Richard, Mohamed, and Jonathan: Thank you very much for your kind words and your contributions to this post.

Paul: Thank you for your note. Strunk & White's The Elements of Style has been a tremendously important influence in my professional life.

Jeff said...

Eisenhower had to assert to U.S. citizens that HE was in control of his troops, not Monty or anyone else. He and the rest of the command staff had to face criticism that U.S. troops were being commanded by non U.S. personell.

So, I don't particularly see Eisenhower as being brave in writing that he withdrew his troops. What he was trying to say is that he didn't wait for anyone else to withdraw THE troops.

BTW, they weren't Eisenhower's troops, they were the President's. So, again, his use of the possessive is an attempt to reinforce his affect on public opinion about who was in control of the troops.

Mahesh said...

Eisenhower was subtly mounting his campaign for the presidency of the United States. ;)

Another book I would recommend to the readers and Cary would do the same is "The Minto Pyramid Principle". My writing has improved significantly after that.


Joel Garry said...

I came into this backwards from Richard Foote's blog. Very interesting, I learned some new things. I was trained as a research scientist, and my initial reaction was similar to Pete Scott's - as I remembered it, the passive voice was the scientific way. On googling the subject, I see I'm way out of date.

But of course, I have to point you at the top of my search results, , which point out how bad it is to use things like "I believe" "do you see what I mean..." :-O

Can't win, can't break even, can't leave the game.

word: jnjzxnuc

Cary Millsap said...


Good article. I've found myself wanting to say, "I believe..." and I've tried to stop just about every instance of it. First, of course "I believe" whatever it is I'm writing, so adding the words is just redundant. But, second—and I think this is the point of the UNC article—if I simply "believe" something, but I don't have the evidence to support it, then the statement probably has no place in my document.

I will say that, to my mind, there's at least one valid exception: when the whole point is to say that I believe something but cannot [yet] prove it, as an appeal to others who might be able to help.

Unknown said...

Having edited technical documents, why do people always fall back on the passive voice as if it means something? Case in point, "A log will be generated." Generated by whom or what? A process? A server? A bunch of gnomes?

I applaud you for advocating use of the active voice for precision!

Cary Millsap said...

j: Exactly!