Trial Transcripts: Since Rabia is out of the country at the moment, and since there are transcripts that are overdue for release, Rabia asked if I could post the next batch of transcripts:
According to Serial folklore, the reason Adnan became a suspect in Hae’s murder was all due to an anonymous call made the day after Hae’s murder was announced in the media. On February 12, 1999, at 3:19 p.m., Detective Massey received a phone call from an “Asian Male 18-21 years old[ ] who advised investigators should concentrate on the victim’s boyfriend.” It was only after this phone call that the police began to zero in on Adnan as a suspect.
Or so the story goes. The police files, however, tell a different story. They indicate that Adnan was already a suspect before the anonymous call ever came in. In fact, the police files indicate that Adnan was the only suspect that was ever considered. As of February 11th, the police already seem to have decided that Adnan was responsible for Hae’s murder.
The Curious Case of Mr. A and the February 11th Printout
The earliest indication that the police were investigating Adnan comes from a one-page printout of a motor vehicle database search. The search had pulled up registration details for Adnan’s Honda Accord, and the time stamp and terminal ID on the printout show that it was generated on a a computer in the Woodlawn police station on February 11, 1999, a little after 8:00 pm.
On the same evening that the printout was made, a man had walked into the Woodlawn police station to report that he had witnessed something he believed to be connected to Hae’s murder. The man — “Mr. A” — told the police that had seen suspicious activity while in Leakin Park, and that he “had heard of the recovery of a woman’s body in Leakin Park on the TV news and wanted to help.” The officers at the Woodlawn station called the Homicide Unit down in Baltimore, since the city now had jurisdiction over the offense. Detective Ritz and Sergeant Lehmann drove out to Woodlawn to talk to Mr. A, arriving at the station at 9:45 pm. According to Lehmann’s report of the interview, Mr. A told them that he had observed
a [young black male] driving a light colored automobile while in Leakin Park . . . acting suspicious near the concrete barriers blocking southbound traffic onto Weatheredsville Road from Windsor Mill Road, which is approximately a mile from the site of the victim’s recovery.
Lehmann’s report then summarily concluded that “investigators believe that this observance is not connected to the murder of Hae Lee.” However, Lehmann provided no indication whatsoever as to why the investigators believed that Mr. A’s report was not connected to Hae’s murder. Lehmann’s summary report also omits a number of crucial details about the interview with Mr. A, including (1) the date and time that Mr. A had seen this “suspicious” activity, (2) what the suspicious activity consisted of, and why exactly it had struck him as suspicious, or (3) why he thought that this event might have been connected to Hae’s murder.
As a result, the report itself provides no basis from which Mr. A’s story can be discredited, or from which it can be concluded that it is irrelevant to Hae’s murder. Based on the timing of his report, Mr. A probably went to the police station immediately after learning that Hae’s body had been found in Leakin Park. Although Hae’s body had been found on Tuesday, February 9th, it was not reported until the evening of Thursday, February 11th, when the story was covered in the evening news (such as in the media segment shown here). The timing of Mr. A’s report would therefore suggest that he had high level of confidence in the importance of what he had seen. After seeing a TV segment on a body being found in Leakin Park, he remembered the incident that he had observed, and — without delay — he immediately drove down to the police station to report what he had seen.
So what exactly caused the officers to reject his report as unrelated to the murder? The location of the suspicious activity that Mr. A reported cannot explain the officers’ dismissal of his story. Although it took place a mile away from where Hae’s body was found, it was a location that would have been an ideal spot to check out, for anyone who might be scouting out Leakin Park for a place to bury a body:
Google Street View of “southbound [ ] onto Weatheredsville Road from Windsor Mill Road.” Note that the concrete barriers referenced in Mr. A’s report were subsequently replaced with the yellow gate shown in this image.
Although Windsor Mill, like N. Franklintown, is a busy by-pass road with significant traffic volume, Weatheredsville is blocked off from the public, and has no traffic at all. Although there were concrete barriers in place to bar vehicles from getting through, if someone could have gotten around the barriers they would have been rewarded with access to a highly secluded, half-mile stretch of road in the middle of Leakin Park, with no traffic whatsoever. A perfect location for body burying — or at least as perfect as it is going to get in Leakin Park. Perfect, that is, if you can get around the concrete barriers to access the road. (And perhaps that is precisely what Mr. A saw — someone trying to get around the concrete barriers could also explain why Mr. A thought the young man he saw was “behaving in a suspicious manner.”)
If the location of the activity can’t explain why the police rejected Mr. A’s witness report, then perhaps instead it was the date on which Mr. A observed the activity that caused the police to dismiss his story. But if so, why would Sergeant Lehmann not have included that detail in his report? Besides, at this point, the police had no idea if Hae’s body had been the park for one week or four — if Mr. A had observed this any time between January 13th and February 8th, then it could have been related to the murder, and should have at least been investigated further before being dismissed out of hand.
So what was it then? Well, Lehmann’s sparse report does include two details from Mr. A’s statement which could explain why investigators rejected his story as irrelevant: Mr. A saw a “black male” who was driving a “light colored” car. If the police had already established that a Pakistani male driving a dark colored car was responsible for the crime, then Mr. A’s report would be irrelevant to the investigation, and could safely be dismissed as an unrelated occurrence.
The printout from the MVA database indicates that Mr. A’s description of the car was, at least, one of the reasons that his report was rejected:
Printout from MVA database, showing that on February 11, 1999 at 8:06 p.m., someone using in the Woodlawn precinct pulled up the vehicle registration records for Adnan’s 87 Honda Accord.
The scan of this document makes it a little hard to read, but the second line provides the VIN for the car — which, with a VIN decoder, the officers could use to look up the color of Adnan’s car.
While it isn’t odd that the police would have had a copy of Adnan’s vehicle registration records in their case file, what is odd is the date and time on which this registration record was obtained: February 11, 1999, at 8:06 pm. The scanner quality of the printout is not great, but a comparison between this printout and another printout — which uses an identical font and was also created on February 11, 1999 — shows that the printout of the Honda Accord’s registration details had in fact been done on February 11th as well. (The blue pixels represent the tag printout, and are overlaid on the date taken fro the sample printout.)
(And yes — for the record, I did compare the numbers on the tag printout with every other possible combination. The result was that the only number combination that matches is “021199.”)
From reviewing the rest of the police files, the time stamp on this printout is further evidence that the search of Adnan’s vehicle records was not something that had been pulled up as a routine matter. Every other printout in Adnan’s case file was made during normal business hours, from 9am to 5pm — so why, then, was a police officer printing out Adnan’s vehicle registration at 8pm on a Thursday evening? This indicates that the printout was generated in response to a particular event, and that an officer had been prompted at that time to look up the information, rather than for the purpose of making routine updates to a case file.
An event such as, for example, Mr. A’s walk-in report at the police station. We do not know what time he arrived at the Woodlawn stationhouse, but if his police report was prompted by an early evening news segment about Hae’s body being discovered in Leakin Park, then sometime around 8 p.m. would make sense. Although Lehmann’s report indicates that he and Ritz did not arrive at the Woodlawn station until 9:45 p.m., Mr. A could easily have arrived into the station a couple hours earlier in the evening — particularly if the Baltimore County officers had interviewed Mr. A themselves first before notifying the Homicide Unit. It would have also taken Lehmann and Ritz some time before they could have arrived at the Woodlawn station to interview Mr. A themselves, particularly as Ritz and MacGillivary appear to have been working a day shift at that time, and probably had to be contacted during their off hours to be notified of the witness.
Unfortunately for the Detective Ritz, the witness report turned out to be a dud. Mr. A saw a young black guy in a light car, but their suspect was a young Pakistani guy in a brown or black car. Ergo, what Mr. A saw could not have been related to Hae’s murder. Yes, Hae’s car could be described as “light colored” — but at this stage of the investigation, the detectives had never heard of the “trunk pop” story. As later documents indicate, the police seemed to have believed that it was Adnan’s car, not Hae’s car, that had been used to transport the body. (See, i.e., the vehicle processing report for Adnan’s car, noting that “Suspect along with witness used the vehicle to aid in the transport . . . of the victim’s body in the trunk section of auto.”). This would explain, then, the easy dismissal of Mr. A’s account: the investigators didn’t need to hear anything more from him, because what he had seen in Leakin Park had been the wrong car driven by someone of the wrong race.
Without knowing what time Mr. A arrived at the station, the connection between the tag printout’s timestamp and Mr. A’s police report cannot be conclusively shown. (Although a records request directed to the Baltimore County policy might go a long way to resolving the question.) Still, regardless of the exact timing, the fact remains that a police officer was investigating the details of Adnan’s car at 8 p.m. on a Thursday, only two days after Hae’s body was found — and why would the officer have done so, if Adnan was not a suspect? But why would Adnan have been a suspect on February 11th, when no anonymous phone call had yet been made implicating him in Hae’s murder?
One theory: although unpreserved in the documentary record, a request for Adnan’s cellphone records had already been made, which had alerted investigators to the existence of the two Leakin Park phone calls.
Following the Subpoena Trail, or, How the Investigators Came to Be in Possession of Adnan’s Cellphone Records Without Ever Issuing a Subpoena
According to Serial, the detectives did not subpoena Adnan’s cellphone records until February 18, 1999, over a week after Hae’s body had been found in Leakin Park. This version of events, as told in the podcast, matches the “official” story of the investigation — that is, it matches the story of the investigation that was disclosed to the defense counsel.
But the police files produced in response to the 2014 MPIA request show that the first subpoena for cellphone records was not issued on February 18th, as the documents given to defense counsel claimed, but had instead been issued two days previously, on February 16th. Moreover, the files also show that, at the time this first subpoena was issued on February 16th, the police were — inexplicably — already in possession of at least some of Adnan’s cellphone records, despite the absence of any documentation concerning the source of this information.
In the state’s production of documents to Adnan’s defense counsel, the state included two police reports concerning subpoenas issued to Adnan’s cellphone provider. Those reports indicate that the first subpoena for cellphone records had been issued on February 18, 1999. In a Progress Report dated 2/18/99, MacGillivary wrote that:
On 18 February 1999, your investigator along with Detective William Ritz obtained a subpoena for the cell phone records of one Adnan Syed telephone # 410-253-9023 from Sgt. Michael Cannon H.l.D.T.F. The subpoena will be delivered on 19 February 1999 to Bell Atlantic Mobile Security, Cockeysville, Maryland.
In a second Progress Report, dated 2/20/99, MacGillivary wrote that a (second?) subpoena had also been served two days later, on February 20th:
In furtherance of the above captioned investigation, on 20 February 1999 at 1115 hours, this writer faxed a subpoena and court order to AT&T Wireless Communications, located 801 North Point Parkway, West Palm Beach, Florida 33407, (fax# 1-888-938-4715). The court orders request that AT&T Wireless provide this investigator with the cell site locations of calls that were made from telephone number (443) 253-9023 [Adnan’s cellphone number] during the time span of January 1, 1999 to present. This writer also asked that a directory of cell site locations associated with the requested information be provided for proper interpretation.
These Progress Reports were inaccurate and misleading, however. They do not indicate (and in fact, seem to be obfuscating) the existence of the prior 2/16/99 subpoena, in response to which AT&T had produced on February 17th Adnan’s subscriber information and call records:
Moreover, although the undisclosed 2/16/99 subpoena is the first documented request for cellphone records that is contained in the police file, the wording of the subpoena indicates that it was not the first ever request for cellphone records that investigators had made. Note the section of the subpoena in bold below:
You are therefore directed this 16th day of February, 1999, to furnish the name(s) address(s) for the following telephone number and (13) cell site locations, from January 1999 to present.
The investigators wanted addresses for “13 cell site locations.” This would indicate that, at the time the 2/16/99 subpoena was issued, the investigators already had information concerning Adnan’s cellphone records, including tower data, because on the day of Hae’s murder, Adnan’s cellphone made calls on 13 separate antennas. But how did the investigators know that on February 16th, when no documentation exists indicating a request had already been made to AT&T at that time, or that AT&T had produced documents in response to such a request?
To me, it looks a lot like someone in the Homicide Unit had already made some unofficial requests for cellphone data, and the phone company had obliged, even in lieu of a formal subpoena requesting production of that data. Perhaps it was something done over the phone, or through a contact in another government agency, but — somehow — the investigators seem to have known only about the significance of Adnan’s cellphone location data only few days after Hae’s body had been found.
The records in the MPIA files also suggest that investigators made an effort to camouflage the fact that previous requests for cellphone location data had been made. Let’s take a look back at the Progress Reports that were disclosed to defense counsel. In the first one, the 2/18/99 Progress Report, MacGillivary wrote, “On 18 February 1999, your investigator along with Detective William Ritz obtained a subpoena for the cell phone records of one Adnan Syed.” This is not an accurate statement. It also creates the (false) impression that the 2/18/99 subpoena was the first records request submitted by investigators. In fact, not only was the 2/18/99 subpoena not the first request for cellphone records that had been submitted by the investigation, it was not even a request for Adnan’s cellphone records at all! Instead, it was a request for the subscriber information for the people that Adnan’s cellphone had called.
The 2/18/99 subpoena. Numbers listed correspond to outgoing calls made by Adnan’s cellphone on January 12 – 14, 1999.
Note: MacGillivary seems to have been calling in a favor from someone with the Drug Enforcement Administration, because the subpoena was issued by the DEA, and not through the channels used by Maryland state agencies. The information requested in that subpoena, however, is clearly not Adnan’s cellphone records, as the 2/18/99 Progress Report describes — instead, it is a subpoena for the subscriber data of almost everyone that was called by Adnan’s phone on January 12th, 13th, and 14th. (Not all numbers called by Adnan’s phone are included in this subpoena, however. Intriguingly, Jenn’s home phone number is not included in this subpoena. Police notes also show that the investigators already knew who Jenn’s home phone number belonged to prior to the 2/18/99 subpoena.)
So why did MacGillivary incorrectly describe the 2/18/99 subpoena as “a subpoena for the cell phone records of one Adnan Syed”? It could be sloppy paperwork; nothing more than the result of an oversight, due to confusion with the February 16th subpoena, which did request Adnan’s cellphone records. But MacGillivary should have known that, as of February 17, 1999, AT&T had already faxed the Homicide Unit the subscriber information and call logs from Adnan’s cellphone — the 2/18/99 subpoena could not have been for Adnan’s cellphone records, when those had been sent over the day before. Here is the front page of the cellphone records sent to the Homicide Unit by AT&T on February 17th:
In addition to the subscriber information above, the fax records sent by AT&T on February 17th also included a record of all outgoing and incoming calls made from Adnan’s cellphone from January 11th through February 16th — but as I’ve discussed in a previous post, the location data for those records had been redacted. Now, it is possible that AT&T was responsible for those redactions, because on February 20th, a third subpoena is issued — this time with a court order compelling AT&T’s compliance. (Note: The fax record for the 2/20/99 subpoena show that it was sent at 11:15 a.m. on February 20th, which means this was the same subpoena that MacGillivary references in the 2/20/99 Progress Report.) This suggests that a court order was in fact needed to obtain the location data, and AT&T may have redacted the February 17th records.
But here is the problem with that scenario: if AT&T was responsible for the redaction of the starting and ending tower location data in the 2/17/99 records, then why does MacGillivary’s fax cover to the 2/20/99 subpoena indicate that he already possessed the cell site location data?
“Please include a cell site directory that corresponds with the sites listed.” In other words: MacGillivary already possessed a list of cell sites that Adnan’s phone made calls on. The Homicide Unit had already been given cell site information as of February 20th (although they seem to have lacked the directory which provided the addresses for those cell sites). But the only cell site records in the police file that pre-date February 20th look like this:
Although the redaction in blue (on the left, redacting phone numbers) was done by me, the redaction in black (on the right, redacting cell sites) was done by hand on a paper copy of the document; no unredacted version exists in the files. Yes, this particular copy of the cell records may have already been redacted when AT&T faxed it over on 2/17/99 — but if so, then where is the unredacted cell site location data that the investigators did have possession of on February 20th, as shown by both MacGillivary’s 2/20/99 fax cover and the 2/16/99 subpoena? Whether or not AT&T redacted this particular document, there should be, somewhere, a document that contains unredacted cell site data for at least some of Adnan’s cellphone records, which pre-dates February 20th. So where is it?
This is potentially a Big Deal. If the unredacted version of the 2/17/99 fax from AT&T showed, for instance, that one of the “Leakin Park calls” had originated on a tower miles away from Leakin Park, that would be exculpatory. It would show, at a minimum, that there would be no reason to believe the cellphone was in Leakin Park, as opposed to somewhere closer to another tower the call had connected through.
The Story of Adrian Syedd’s Traffic Violation, and Why It May Indicate That Investigators Had Adnan’s Cellphone Records Before Any Formal Request Was Made
In addition to the subpoenas themselves, there is another piece of evidence that suggests investigators had obtained Adnan’s cellphone records before any documented request for that data had been submitted. That evidence comes from the Maryland court records, which show that Adnan was issued a citation for a traffic violation on February 15, 1999. However, circumstantial evidence from the police files indicates that this traffic stop may in fact have been connected to the murder investigation, and, more specifically, to the investigators’ efforts to obtain his cellphone records.
First, on February 14, 1999, the police checked Adnan’s records on the MVA database once again. This time, rather than pulling up information based on Adnan’s vehicle registration, the police pulled up Adnan’s records based on his driver’s license:
On its own, this search is easily explainable based on the fact that, a couple days before, an anonymous call and been made implicating Adnan in Hae’s murder. It is not odd that Adnan’s driving history would have been pulled at some point, and added to the case file. The timing of this search, however, suggests that the police had a specific motivation for searching Adnan’s MVA records on that date, as the following day — February 15th — Adnan was pulled over for a seatbelt violation:
I wouldn’t have questioned whether this routine traffic stop was, in fact, simply a routine traffic stop, if it were not for one glaring error in the record: the police officer who made the traffic stop issued the citation not to “Adnan Syed,” but to “Adrian Syed” instead. Although the license plate indicates that this was in fact Adnan’s car that had been pulled over, both the street address an Adnan’s name are incorrect.
In regards to the misnomer, this was not the first time that this particular error has been made. It was the second. The first time was in Adnan’s cellphone records, in which AT&T had mistakenly recorded his name as Adrian Syedd:
The source of the error, in the context of the AT&T billing records, could be a result of AT&T’s billing software mistakenly misreading the first “n” in Adnan as “ri.” (This appears to be a potential problem that could still occur today — if I search my notes for the term ‘Adrian,’ I pull up about a hundred instances in which ‘Adnan’ is typed in a screenclipped image.) The error makes less sense in the context of a traffic citation, but maybe it could happen. What are the odds, though, that when Adnan was pulled over for a traffic violation just two days after becoming a suspect in a murder investigation, the police officer who issued the citation would make the exact same odd spelling error that was made on Adnan’s cellphone records — which the police would (officially) obtain from AT&T two days later, on February 17?
Instead of a coincidence, though, it could be an indication of a connection between the cellphone records and the traffic stop. The error in the cellphone records could potentially have been both motivation for the traffic stop, and the cause of the naming error — if, say, the traffic stop was used as a way to confirm Adnan’s possession of the cellphone associated with the records that had been provided by AT&T.
If the investigators had somehow obtained Adnan’s cellphone data through the use of “unofficial channels,” then the fact that the subscriber information for those records was not listed under Adnan’s real name would have been the cause of some concern among the investigators. Because the records from AT&T were for someone named “Adrian Syedd,” and not for someone named “Adnan Syed,” the investigators may have needed confirmation that the subscriber data they had obtained did, in fact, belong to the suspect in their murder investigation. But, since this hypothetical cellphone data would not have come from official sources (no subpoenas had been issued yet), how could the investigators have proven that the cellphone records were for Adnan’s phone, without using “official” channels? In other words, how could they unofficially verify the accuracy of their unofficial records?
Pulling Adnan over for a traffic stop, and having someone make a call to his cellphone while that officer watched, would be one way of doing that. That would allow investigators to confirm that the cellphone records they had pulled did belong to the cellphone owned by their suspect (instead of by some unknown relative named Adrian). In support of this theory, there was in fact a brief incoming call made to Adnan’s phone at 12:11 p.m. — the same time listed as the time of the traffic citation.
If the scenario outlined here is what actually occurred, then the investigators’ testing would appear to have been a successful one, as the following day, on February 16, an official subpoena for Adnan’s cellphone records was issued by the grand jury. The investigators appear to have obtained whatever information they needed to proceed with an official request for Adnan’s cellphone records. The officer who made the traffic stop may have slipped up, however — and rather than writing down the traffic offender’s name as it was shown on his driver’s license, he wrote down the false name displayed on the cellphone records he was attempting to verify.
What All of This May Mean
From a review of the subpoenas and cell record data, it looks like: (1) the investigators had cell record data of an undisclosed nature and from an undisclosed source; (2) the investigators had identified Adnan as their suspect before any (disclosed) evidence implicating him in the murder had been uncovered; and (3) MacGillivary had contacts with the Drug Enforcement Administration — an agency that was an early trailblazer in the use of cellphone location data as an investigative tool for law enforcement.
There are a few different things this could mean, but the question all of that raises for me is this: could Adnan have been identified as a suspect as a result of a warantless cell tower dump on L689?